- Email: [email protected]
Subclinical vascular disease and the risk of parkinsonism: The Rotterdam Study
Parkinsonism and Related Disorders xxx (2017) 1e6
Contents lists available at ScienceDirect
Parkinsonism and Related Disorders journal homepage: www.elsevier.com/locate/parkreldis
Subclinical vascular disease and the risk of parkinsonism: The Rotterdam Study Vanja Vlasov a, Sirwan K.L. Darweesh a, b, c, Bruno H. Stricker a, d, Oscar H. Franco a, M.Kamran Ikram a, b, Maryam Kavousi a, Daniel Bos a, c, e, Caroline C.W. Klaver a, f, M.Arfan Ikram a, b, e, * a
Department of Epidemiology, Erasmus MC University Medical Center Rotterdam, Rotterdam, The Netherlands Department of Neurology, Erasmus MC University Medical Center Rotterdam, Rotterdam, The Netherlands Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, MA, USA d Inspectorate for Health Care, 2595 AN, The Hague, The Netherlands e Department of Radiology, Erasmus MC University Medical Center Rotterdam, Rotterdam, The Netherlands f Department of Ophthalmology, Erasmus MC University Medical Center Rotterdam, Rotterdam, The Netherlands b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 10 February 2017 Received in revised form 3 June 2017 Accepted 27 June 2017
Background: Parkinsonism is a common neurodegenerative syndrome in middle-aged and elderly persons. The etiology is multifactorial with a possible vascular contribution, but this has not been comprehensively studied. Objective: To determine whether selected markers of subclinical vascular pathology are associated with the risk of all-cause parkinsonism in the general population. Methods: We assessed a range of markers of subclinical vascular pathology (ankle-brachial index, carotid plaques and intima media thickness, retinal arteriolar and venular calibers) in 6199 persons from the population-based Rotterdam Study, who were free of parkinsonism and dementia at baseline. We followed these persons up till onset of parkinsonism, dementia, and death for 89,387 person-years until January 1, 2013. Hazard ratios (HRs) for all-cause parkinsonism and separately for Parkinson disease (PD) versus non-PD were estimated from competing risk regression models adjusting for potential confounders. Results: During follow-up, we identified 211 cases of parkinsonism (110 had PD). None of the five markers of subclinical pathology was associated with all-cause parkinsonism. Only low ankle-brachial index was associated with a higher risk of non-PD parkinsonism (HR ¼ 0.79, 95%CI: 0.68e0.92), but not with the risk of PD. Conclusion: We did not find a consistent pattern of associations between systemic vascular pathology markers with parkinsonism, suggesting that the potential involvement of vascular pathology is not prominent or needs further evaluation in studies with an even larger sample size. © 2017 Elsevier Ltd. All rights reserved.
Keywords: Parkinsonism Vascular pathology Risk factors Etiology
1. Introduction Parkinsonism, with its most frequent subtype Parkinson disease (PD), is a common chronic neurodegenerative syndrome in the elderly [1,2]. Although many genetic and environmental factors have been implicated in the multi-factorial etiology of parkinsonism [1], there is still much uncertainty about the exact
* Corresponding author. Department of Epidemiology, Erasmus MC University Medical Center, Dr. Molewaterplein 50, 3015 GE, Rotterdam, The Netherlands. E-mail address: [email protected] (M.Arfan Ikram).
mechanisms underlying neuronal cell loss in patients with Parkinson disease and non-PD parkinsonism. There is a scarcity of knowledge on early markers and potential neuroprotective therapies in early stages of parkinsonism, and patients are typically already in an advanced pathological disease stage when the diagnosis is made [1]. Therefore, there is a need for novel insight into factors associated with all-cause parkinsonism. Markers of subclinical vascular pathology are strongly implicated in the etiology of the two common neurological syndromes, stroke and dementia [3]. Furthermore, a high prevalence of lacunar infarcts have been reported in the cerebral white matter and basal
http://dx.doi.org/10.1016/j.parkreldis.2017.06.022 1353-8020/© 2017 Elsevier Ltd. All rights reserved.
Please cite this article in press as: V. Vlasov, et al., Subclinical vascular disease and the risk of parkinsonism: The Rotterdam Study, Parkinsonism and Related Disorders (2017), http://dx.doi.org/10.1016/j.parkreldis.2017.06.022
2
V. Vlasov et al. / Parkinsonism and Related Disorders xxx (2017) 1e6
ganglia of patients with parkinsonism [4], which indicates a potential vascular etiology. During the course of Alzheimer's disease 25% of patients develop parkinsonism [5], whereas approximately 30% of patients with PD are eventually diagnosed with dementia [6]. However, in spite of an overlap in clinical and pathological findings between primary parkinsonism (such as PD) and dementia or stroke, the role of vascular pathology in the etiology of parkinsonism syndromes remains unclear. Vascular pathology can be non-invasively measured both in large systemic vessels as well as in microvasculature locally [7]. Commonly used, non-invasive measurements of vascular pathology of the large vessels include ultrasound measures of carotid arteries (intima media thickness and plaques) and ankle-brachial index (ABI) [8,9]. Furthermore, structural changes in the retinal vasculature have long been recognized as an important marker of microvascular pathology [10]. These non-invasive measures provide an insight into the condition of the large and small blood vessels. We hypothesized that subclinical vascular pathology is associated with an increased risk of parkinsonism in the general population. We used noninvasive measurements of subclinical vascular disease to test our hypothesis in a large, population-based cohort study. 2. Methods 2.1. Setting and study population This study was conducted as part of the Rotterdam Study, a community-based prospective cohort study in elderly adults [11]. There were only two selection criteria: persons had to live in the well-defined Ommoord district of Rotterdam in 1990, and they had to be aged 55 years or older. Out of 10,215 eligible individuals, 7983 agreed to participate and provided informed consent. All participants were interviewed and underwent extensive physical examination at the research center, including screening for parkinsonian signs and markers of vascular pathology at baseline (1990e1993) and at four follow-up rounds (1993e1995, 1997e1999, 2002e2004, and 2009e2011). For the current study we excluded participants with prevalent parkinsonism (n ¼ 126), incomplete neurological screening (n ¼ 1401) at baseline, and participants with missing information regarding two or more vascular pathology markers (n ¼ 257), leaving 6199 subjects at risk to develop any kind of parkinsonism during the follow-up period. Similarly with other population-based studies, the Rotterdam Study overall age- and sex-adjusted incidence rate of Parkinson Disease is 1.5 per 1000 person-years [2], while age-specific incidence rates are somewhat higher than in community based studies [12]. The Rotterdam Study has been approved by the medical ethics committee according to the Population Screening Act: Rotterdam Study, executed by the Ministry of Health, Welfare and Sports of the Netherlands. All participants provided written informed consent. 2.2. Assessment of subclinical vascular disease Recently, CT and MRI assessments were implemented in the Rotterdam Study; however, we lacked sufficient follow-up for incident parkinsonism to incorporate these measurements in our present analyses. Subclinical vascular pathology was assessed with four different measures: ankle-brachial index, carotid intimamedia thickness and the presence of plaques in the carotid arteries assessed using ultrasonography and retinal vessels diameters.
2.3. Ankle-brachial index We assessed ankle-brachial index (ABI) by measuring systolic blood pressure (SBP) in all extremities. The SBP was measured twice in both arms and at both ankles (posterior tibial artery) in each participant in the supine position with an 8-MHx continuouswave Doppler device (Huntleigh D500, Hunteleigh Technology) which was attached to a standard random-zero sphygmomanometer [8]. To obtain the ABI we calculated the ratio of the systolic blood pressure at the ankle and the systolic blood pressure at the arm per side. ABI values higher than 1.50 were not included in the study, because that values are considered irrationally high due to calcification of the blood vessels and their inability to properly compress. We used the lowest value of the two sides for the analyses [8].
2.4. Carotid number of plaques and intima media thickness Using B-mode ultrasonography we evaluated both carotid arteries for the presence of plaques and determined the carotid intima media thickness (cIMT) [9]. The presence of a plaque was defined as a focal widening relative to bordering segments, with protrusion into the lumen consisting of either echolucent or echogenic areas. We assessed the presence of plaques at both sides in the common carotid artery, carotid bifurcation and proximal part of the internal carotid artery. The total amount of plaques was reflected in a weighted plaque score that ranged from 0 to 6 [13]. The average of the maximum IMT was registered at both the near- and far wall of the common carotid artery for assessing the cIMT.
2.5. Retinal vessel measurements At the baseline ophthalmic examination, simultaneous stereoscopic fundus color transparencies were taken centered on the optic disk (20 field, Topcon Optical Company, Tokyo, Japan) after pharmacologic mydriasis and were digitized with a high-resolution scanner (Nikon LS-4000, Nikon Corporation, Japan) [14]. Evaluated measures for arteriolar and venular calibers were based on improved Parr-Hubbard formulas and were corrected for magnification changes due to refractive errors of the eye [14]. Since each eye has a different magnification in case of refractive changes, summary vessel measures were additionally adjusted for the refraction and corneal curvature with Littmann's formula to estimate absolute measures [14].
2.6. Other measurements We assessed cardiovascular risk factors by interview, physical examination and blood sampling [11]. Blood pressure was calculated as the mean of two consecutive measurements with a random-zero sphygmomanometer at the right brachial artery while the patient was in a sitting position. Serum total cholesterol was determined by an automated enzymatic procedure in a non-fasting blood sample. Serum high density lipoprotein (HDL) cholesterol was measured after precipitation of the non-HDL fraction with phosphotungstate-magnesium. Height and weight were measured and body mass index (kg/m2) was calculated. Non fasting serum glucose levels were measured using the glucose hexokinase method. Smoking habits of participants were categorized as “ever” or “never”. Uric acid was determined with a Kone Diagnostica reagent kit and a Kone auto-analyzer from blood samples of participants.
Please cite this article in press as: V. Vlasov, et al., Subclinical vascular disease and the risk of parkinsonism: The Rotterdam Study, Parkinsonism and Related Disorders (2017), http://dx.doi.org/10.1016/j.parkreldis.2017.06.022
V. Vlasov et al. / Parkinsonism and Related Disorders xxx (2017) 1e6
2.7. Ascertainment of parkinsonism and parkinson disease A detailed description of parkinsonism and Parkinson disease assessment methods has previously been published [2] and is summarized in Supplement 1. In short, we used four overlapping modalities to screen for potential parkinsonism during follow-up: in-person screening (on average every 4 years), in-person interviews, use of antiparkinsonian medication, and alerts from continuous monitoring of clinical records. Of all persons who screened positive in any of these methods, complete medical records (including letters from medical records of specialists and general practitioners) were studied and case reports were drawn up covering all potentially relevant information to establish presence and cause of parkinsonism. These case reports were evaluated by a panel led by an experienced neurologist. Parkinson disease was only diagnosed after exclusion of secondary causes, and medical records of all incident parkinsonism cases (both PD and non-PD) continued to be scrutinized until the end of the study period for new information that could lead to a revision of the diagnosis. Given the substantial overlap between the four detection methods we considered persons who were not screened in-person during one of the follow-up rounds still at risk for parkinsonism and Parkinson disease. Person-time at risk of incident parkinsonism ended at onset of parkinsonism, death, or January 1, 2013. Person-time at risk of incident Parkinson disease ended at onset of parkinsonism, incident dementia (date of clinical diagnosis), stroke, death, or January 1, 2013. 2.8. Statistical analysis Baseline characteristics are presented as frequencies with percentages for categorical variables or means (standard deviation [SD]) for continuous variables. Potential differences in baseline characteristics between included and excluded individuals were explored using independent t-tests for continuous variables and chi-square tests for categorical variables (Supplement 2). The sample size and power calculations were calculated a priori and presented in Supplement 3. The association of measures of subclinical vascular disease with risk of parkinsonism was investigated using competing risk regression models. We analyzed each risk factor separately. All continuous vascular pathology markers were transformed into zscores to facilitate direct comparison of effect sizes. We performed similar analyses for the risk of Parkinson disease and non-PD parkinsonism. For non-PD parkinsonism, we considered PD and death to be competing events; for PD, we considered non-PD parkinsonism, death, stroke and dementia as a competing events. All analyses were adjusted for age and sex (model 1). Additionally, we adjusted our analyses for systolic blood pressure, diastolic blood pressure, body mass index, total cholesterol level, HDL cholesterol level, glucose level, smoking status and serum uric acid (model 2). In all analyses involving retinal vessel diameters we adjusted these for each other (venular diameters for arterial diameters and vice versa). To explore possible effect modification in our analysis we separately added interaction terms of sex and age with markers of subclinical vascular pathology (i.e. sex*ABI; age*ABI) and conducted stratified analyses on sex and on age (using mean age as the threshold). To investigate whether association of ABI and non-PD parkinsonism were driven by drug-induced parkinsonism, we repeated analyses for association after excluding 20 medicationinduced cases. In order to assess subclinical vascular disease measures as prodromal markers instead of risk factors, we repeated main analyses after making cut offs in a follow-up time on two years, five years and ten years (Supplement 4). We conducted sensitivity analysis on
3
estimates of the association between subclinical vascular markers and all-cause parkinsonism after censoring cases of unspecified parkinsonism. Moreover, we stratified Parkinson disease patients into groups with and without subclinical vascular pathology, and compare them according to the age of onset of the disease, UPDRSIII score, treatment and a mortality rate from the date of onset of parkinsonism until the date of censoring. We had complete data on 95% of independent variables. Missing values were imputed by multiple imputation (n ¼ 5), based on gender, age and all others covariables. IBM SPSS Statistics version 21.0 (IBM Corp, Armonk, NY, USA) and R version 3.2.5 (R Foundation for Statistical Computing, Vienna, Austria) were used for statistical analyses. 3. Results Baseline characteristics of the study population are presented in Table 1. The mean age at baseline of the eligible 6199 people was 68.5 years and 58.7% of the participants was female. A comparison between included and excluded individuals demonstrated that excluded individuals were older, had a significantly higher body mass index, higher ABI, larger cIMT, and higher serum glucose level (Supplement 2. Table 1.). During a total follow-up of 89,387 person-years, [mean (SD), 14.42 (6.5)], we identified 211 patients with parkinsonism, of whom 110 patients were diagnosed with Parkinson disease. An overview of diagnoses is provided in Table 2. Table 3 shows the association between markers of subclinical vascular disease and the risk of all-cause parkinsonism, Parkinson disease and non-PD parkinsonism. Overall, no measure of subclinical vascular disease was associated with the risk of all-cause parkinsonism. Correspondingly, none of the vascular markers in our study were associated with risk of Parkinson disease. We only found a decreased risk for non-PD parkinsonism for participants with higher ABI [HR: 0.79 (95% CI: 0.68e0.92)]. After excluding drug-induced cases, associations of ABI and the non-PD parkinsonism remained almost unchanged [HR: 0.77 (95% CI: 0.63e0.93)]. Furthermore, carotid plaques, cIMT and diameter of retinal vessels were not associated with non-PD parkinsonism. Additional adjustment for cardiovascular risk factors did not substantially alter any of these results (Table 3, Model 2). We found no evidence for statistical interaction by sex or age (all p-interaction>0.73). Additional analyses of subclinical vascular pathology markers as predictors of all-cause parkinsonism over restricted follow-up
Table 1 Baseline characteristics of the study population. Characteristic
Participants (n ¼ 6199)
Age at baseline, years Women, % Smoking, % Body mass index, kg/m2 Total cholesterol, mmol/L HDL cholesterol, mmol/L Systolic blood pressure, mm Hg Diastolic blood pressure, mm Hg Serum glucose, mmol/L Serum uric acid, mmol/L Ankle-brachial index Weighted number of plaques, number Carotid intima media thickness, mm Retinal arteriolar diameter, mm Retinal venular diameter, mm
68.5 ± 8.5 58.7 64.5 26.3 ± 3.7 6.6 ± 1.2 1.3 ± 0.3 139.2 ± 22.2 73.8 ± 11.4 6.8 ± 2.5 321 ± 80.7 1.0 ± 0.2 1.5 ± 1.6 1.0 ± 0.2 146.7 ± 14.3 221.7 ± 20.9
HDL, high-density lipoprotein. Values are mean ± standard deviation for continuous variables, percentage for dichotomous variables.
Please cite this article in press as: V. Vlasov, et al., Subclinical vascular disease and the risk of parkinsonism: The Rotterdam Study, Parkinsonism and Related Disorders (2017), http://dx.doi.org/10.1016/j.parkreldis.2017.06.022
4
V. Vlasov et al. / Parkinsonism and Related Disorders xxx (2017) 1e6
characteristics between the two groups. We cannot exclude that selection bias led to slight underestimations of associations between markers of subclinical vascular pathology and parkinsonism. Moreover, we note that we excluded individuals who were younger than 55 years at baseline, and that this selection could have introduced bias if the association between subclinical vascular disease and incident PD is substantially different in younger individuals. Second, it seems unlikely that unmeasured confounding could account for the null results, but it cannot be completely excluded in observational studies. The role of potential confounders that were assessed was examined in multivariate models, and these were found to have little impact on effect estimates. Third, we used only extracranial measures of subclinical vascular pathology and not the most sensitive markers of subclinical vascular pathology. In the future, community-based studies with adequate follow-up for incident parkinsonism should focus on measures of intracranial vascular pathology imaging, including novel brain-imaging techniques (e.g. calcification imaging on CT, arterial spin labeling on fMRI). Moreover, the vast majority of an unspecified parkinsonism in our study did not have neuroimaging information, those patients could not be classified as PD or other parkinsonism subtypes. Finally, some misclassification of Parkinson Disease cases as other subtypes of parkinsonism might have occurred, since we had no pathological confirmation of the diagnoses. This prospective population-based study has several essential and unique strengths. First, all participants were included and followed up using the same methodology, irrespective of their risk of parkinsonism. Second, the large sample size, long follow-up duration and the carefully collected prospective data on exposure variables as well as on potential confounders minimizes misclassification of baseline vascular pathology profiles of study participants. Furthermore, we actively screened participants for parkinsonism at baseline and only included persons without parkinsonism. Since all assessments of vascular pathology markers took place at baseline, our findings unequivocally reflect the prediagnostic phase of parkinsonism and Parkinson disease. Finally, our near-complete follow-up for dementia and death allowed us to account for events that possibly interfere with the observation of the parkinsonism and Parkinson disease. Previous population-based study has focused mainly on the association between hypertension and the risk of Parkinson disease [15]. However, there is a little prospective data on the role of subclinical vascular disease in the etiology of parkinsonism in the
Table 2 Overview of incident clinical diagnoses of parkinsonism. All-cause PS patients (n ¼ 211) Parkinson disease Unspecified parkinsonism Drug-induced parkinsonism Vascular parkinsonism Multiple system atrophy Lewy body dementia Secondary to dementia Secondary to a tumor Progressive supranuclear palsy
110 (52.1%) 61 (28.9%) 20 (9.5%) 8 (3.8%) 4 (1.9%) 1 (0.5%) 4 (1.9%) 2 (0.9%) 1 (0.5%)
These are clinical diagnoses; no histologic confirmation was obtained. n, number of cases during follow-up.
periods are presented in the supplementary 4, Table 3. In analyses that were restricted on 2-years of follow up, hazard ratios were slightly higher than in the overall analyses. Results from other time-restricted analyses did not differ from the overall analyses. The Supplementary Table 4 shows the results of main analyses after censoring 61 cases with unspecified parkinsonism. No differences were found between the main results and the results after censoring unspecified parkinsonism cases. Parkinson disease patients, who were in the top risk-increasing quartile of at least one subclinical vascular pathology marker, had later onset of a parkinsonism comparing with Parkinson disease patients without vascular pathology. There was no difference regarding to mortality rate, UPDRS-III score and medication use between this two groups.
4. Discussion We did not find an association between markers of subclinical vascular disease and the risk of all-cause parkinsonism or Parkinson disease. For non-PD parkinsonism, we only observed a significant association with lower ABI, but not with any other marker of subclinical vascular disease. Several limitations of this study need to be discussed. First, we excluded participants who were not screened for parkinsonism at baseline and persons with missing data on more than two markers of subclinical vascular pathology. Participants with a missing score were older, on average with lower ABI and with higher cIMT compared to those who were eligible for the study (Supplement 2, Table 1.). No significant differences were found in other measured
Table 3 Associations of markers of subclinical vascular pathology with the risk of parkinsonism. Parkinsonism n/N ¼ 211/6199
Parkinson Disease n/N ¼ 110/6199
Non PD parkinsonism n/N ¼ 101/6199
HR (95% CI)
HR (95% CI)
HR (95% CI)
1.14 0.84 0.86 0.98 0.96
(0.94e1.39) (0.68e1.05) (0.68e1.10) (0.77e1.25) (0.76e1.23)
0.79 (0.68e0.92)* 0.97 (0.80e1.18) 0.97 (0.77e1.22) 0.89 (0.68e1.17) 0.89 (0.67e1.19)
1.08 0.90 0.91 1.04 1.01
(0.87e1.33) (0.72e1.14) (0.69e1.19) (0.82e1.31) (0.78e1.27)
0.76 (0.65e0.90) 0.99 (0.82e1.20) 0.98 (0.78e1.24) 0.96 (0.73e1.26) 0.86 (0.64e1.16)
Per 1-SD increase in:
Model 1
Ankle brachial index Weighted number of plaques Carotid intima media thickness Retinal arteriolar diameter Retinal venular diameter
0.93 0.90 0.91 0.94 0.92
(0.82e1.06) (0.78e1.05) (0.77e1.08) (0.78e1.13) (0.77e1.12)
Model 2 Ankle brachial index Weighted number of plaques Carotid intima media thickness Retinal arteriolar diameter Retinal venular diameter
0.88 0.95 0.94 1.01 0.93
(0.77e1.01) (0.81e1.10) (0.79e1.13) (0.84e1.20) (0.77e1.12)
n, number of cases; N, number of persons at risk; HR, hazard ratio; CI, confidence interval; SD, standard deviation. Model 1: Adjusted for age and sex. Model 2: Adjusted for age, sex, systolic blood pressure, diastolic blood pressure, smoking, body mass index, total cholesterol, HDL cholesterol, glucose, uric acid. *Significant (p < 0.05) data in bold.
Please cite this article in press as: V. Vlasov, et al., Subclinical vascular disease and the risk of parkinsonism: The Rotterdam Study, Parkinsonism and Related Disorders (2017), http://dx.doi.org/10.1016/j.parkreldis.2017.06.022
V. Vlasov et al. / Parkinsonism and Related Disorders xxx (2017) 1e6
general population. Moreover, previous observations were focused on whether vascular pathology contribute to the disease severity in already diagnosed PD patients [16,17] or whether vascular disease co-occurred with Parkinson disease [18,19]. The main novelty of the current study is that we studied markers of several subclinical vascular pathology as risk factors of incident parkinsonism. Remarkably, none of these measures were associated with all-cause parkinsonism and Parkinson disease. Carotid pathology as shown on ultrasound was associated with motor and the cognitive impairment in the PD patients in previous studies [16,17]. Additionally, carotid plaques and higher cIMT have been associated with gait and balance decline over time [20]. Also, patients with vascular parkinsonism have a higher risk for embolic cerebrovascular events [16]. However, another study showed that patients with Parkinson disease had less severe carotid lesions with mostly stable calcified plaques compared to the general population [19]. In this study, we investigated the cIMT and number of plaques as potential risk factors of onset of parkinsonism and Parkinson disease. We observed that ultrasound measurements of carotid artery pathology were not related to a higher risk of all-cause parkinsonism, Parkinson disease or non-PD parkinsonism. A commonly used measurement of vascular pathology in symptomatic and asymptomatic individuals with cardiovascular disease is ABI [8,21]. Ankle brachial index is inversely related to the severity of peripheral arterial disease (PAD), which mean that persons with lower ABI have more expressed pathology of lower extremity arteries [8]. We found that ABI was not related to a higher risk of Parkinson disease but instead with the risk of non-PD parkinsonism. There are several possible explanations for the link between peripheral arterial disease (PAD) and non-PD parkinsonism, such as oxidative stress as same pathophysiological mechanism [22,23] and that vascular pathology could lead to neurodegeneration outside of the nigrostriatal area in non-PD parkinsonism patients [24]. Autonomic dysfunction can be one of the first symptoms in non-PD parkinsonism patients [25]. Moreover, the autonomic nervous system (ANS) has blood supply from peripheral vessels and PAD might lead to neuronal cell loss of ANS. Vascular parkinsonism (VP) can occur after a stroke and studies suggest that atherosclerotic changes have an influence on this type of parkinsonism [26,27]. More than eighty years ago arteriosclerotic parkinsonism was already recognized as a distinct clinical entity by Critchley [28]. While percentage of VP cases among non-PD parkinsonism in our study is low (7.9%), we cannot exclude misclassification of VP as unspecified parkinsonism regarding its much higher percentage in the same substrate (60.4%). Therefore, the association of unspecified parkinsonism with peripheral vascular pathology might warrant further investigation. However, these results should be interpreted with caution because of the lack of associations of other vascular pathology markers and the risk of non-PD parkinsonism. Any impact of multiple testing cannot be ruled out. The diameter of retinal vessels is mainly associated with cerebral small vessel disease (CSVD) [10]. In our study there was no relationship between arteriolar diameters and larger venular diameters with all-cause parkinsonism. This is contrary to two crosssectional studies and one study in patients with CSVD, which revealed that measures of CSVD are associated with the incidence of parkinsonian signs [29e31]. Those studies were using more sensitive measures of CSVD, such as white matter volume, number of lacunes, and cerebral microbleeds. However, their follow-up time was nonexistent or short, and their sample size was smaller than in our study. Similar to our results, one case-control study showed that retinal vessel diameter did not differ between PD patients and healthy controls [32]. Additionally, recent clinicopathological study observed no difference between the incidence of
5
small vessels disease pathology and stroke in PD cases compared with the general population [33]. Different methodology and different types of parameters for CSVD can explain the divergent results from studies on CSVD and parkinsonism risk. In our cohort of middle-aged and elderly adults, potential risk factors for late-onset of parkinsonism were investigated. Still, we cannot exclude the possibility that the subclinical vascular pathology might be associated with young-onset of parkinsonism. In conclusion, we did not find a consistent pattern of associations between measures of subclinical vascular pathology and allcause parkinsonism, thus our report suggest that systemic vascular pathology does not play a large role in the etiology of parkinsonism in the general population. Still, prospective studies that incorporate novel brain-imaging and have adequate sample size and follow-up may detect associations of intracranial vascular pathology with incident parkinsonism. Author contributions Study concept and design: Vanja Vlasov, Sirwan Darweesh, Daniel Bos. Acquisition of data: Arfan Ikram, Sirwan Darweesh, Daniel Bos, Vanja Vlasov. Statistical analysis and interpretation of data: Vanja Vlasov. Drafting the manuscript: Vanja Vlasov. Critical revision of the manuscript for important intellectual content: Arfan Ikram, Kamran Ikram, Bruno Sticker, Oscar Franco, Maryam Kavousi and Caroline Klaver. Study supervision: Sirwan Darweesh, Daniel Bos, Arfan Ikram. Sources of funding This study received support from Stichting ParkinsonFonds. The Rotterdam Study is further supported by the Erasmus MC University Medical Center and Erasmus University Rotterdam, the Netherlands Organization for Scientific Research (NWO), the Netherlands Organization for Health Research and Development (ZonMW), the Research Institute for Diseases in the Elderly (RIDE), the Ministry of Education, Culture and Science, the Ministry of Health, Welfare and Sport, The European Commission (DGXII), the Netherlands Genomics Initiative (NGI) and the Municipality of Rotterdam. The contribution of the inhabitants, general practitioners and pharmacists of the Ommoord district to the Rotterdam Study are gratefully acknowledged. Disclosure statement All authors have approved the final article and report no conflicts of interest. The paper or any part of it has not been published previously and is not under consideration for publication elsewhere. None of the authors report disclosures. Conflicts of interest None. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.parkreldis.2017.06.022. References [1] L.M. de Lau, M.M. Breteler, Epidemiology of Parkinson's disease, Lancet Neurol. 5 (6) (2006) 525e535.
Please cite this article in press as: V. Vlasov, et al., Subclinical vascular disease and the risk of parkinsonism: The Rotterdam Study, Parkinsonism and Related Disorders (2017), http://dx.doi.org/10.1016/j.parkreldis.2017.06.022
6
V. Vlasov et al. / Parkinsonism and Related Disorders xxx (2017) 1e6
[2] S.K. Darweesh, P.J. Koudstaal, B.H. Stricker, A. Hofman, M.A. Ikram, Trends in the incidence of Parkinson disease in the general population: the Rotterdam study, Am. J. Epidemiol. 183 (11) (2016) 1018e1026. [3] H. Gardener, C.B. Wright, T. Rundek, R.L. Sacco, Brain health and shared risk factors for dementia and stroke, Nat. Rev. Neurol. 11 (11) (2015) 651e657. [4] E.E. Smith, J.A. Schneider, J.M. Wardlaw, S.M. Greenberg, Cerebral microinfarcts: the invisible lesions, Lancet Neurol. 11 (3) (2012) 272e282. [5] F. Portet, N. Scarmeas, S. Cosentino, E.P. Helzner, Y. Stern, Extrapyramidal signs before and after the diagnosis of incident Alzheimer disease in a prospective population study, Arch. Neurol. 66 (9) (2009) 1120e1126. [6] D. Aarsland, K. Andersen, J.P. Larsen, A. Lolk, P. Kragh-Sorensen, Prevalence and characteristics of dementia in Parkinson disease: an 8-year prospective study, Arch. Neurol. 60 (3) (2003) 387e392. [7] S. Laurent, M. Briet, P. Boutouyrie, Large and small artery cross-talk and recent morbidity-mortality trials in hypertension, Hypertension 54 (2) (2009) 388e392. [8] F.G. Fowkes, The measurement of atherosclerotic peripheral arterial disease in epidemiological surveys, Int. J. Epidemiol. 17 (2) (1988) 248e254. [9] M.L. Bots, A. Hofman, P.T. De Jong, D.E. Grobbee, Common carotid intimamedia thickness as an indicator of atherosclerosis at other sites of the carotid artery. The Rotterdam Study, Ann. Epidemiol. 6 (2) (1996) 147e153. [10] M.K. Ikram, F.J. De Jong, E.J. Van Dijk, N.D. Prins, A. Hofman, M.M. Breteler, P.T. De Jong, Retinal vessel diameters and cerebral small vessel disease: the Rotterdam Scan Study, Brain 129 (Pt 1) (2006) 182e188. [11] A. Hofman, G.G.O. Brusselle, S.D. Murad, C.M. van Duijn, O.H. Franco, A. Goedegebure, M.A. Ikram, C.C.W. Klaver, T.E.C. Nijsten, R.P. Peeters, B.H.C. Stricker, H.W. Tiemeier, A.G. Uitterlinden, M.W. Vernooij, The Rotterdam Study: 2016 objectives and design update, Eur. J. Epidemiol. 30 (2015) 661e708. [12] L.M. de Lau, P.C. Giesbergen, M.C. de Rijk, A. Hofman, P.J. Koudstaal, M.M. Breteler, Incidence of parkinsonism and Parkinson disease in a general population: the Rotterdam Study, Neurology 63 (7) (2004) 1240e1244. [13] I.M. van der Meer, M.L. Bots, A. Hofman, A.I. del Sol, D.A. van der Kuip, J.C. Witteman, Predictive value of noninvasive measures of atherosclerosis for incident myocardial infarction: the Rotterdam Study, Circulation 109 (9) (2004) 1089e1094. [14] L.D. Hubbard, R.J. Brothers, W.N. King, L.X. Clegg, R. Klein, L.S. Cooper, A.R. Sharrett, M.D. Davis, J. Cai, Methods for evaluation of retinal microvascular abnormalities associated with hypertension/sclerosis in the Atherosclerosis Risk in Communities Study, Ophthalmology 106 (12) (1999) 2269e2280. [15] C. Qiu, G. Hu, M. Kivipelto, T. Laatikainen, R. Antikainen, L. Fratiglioni, P. Jousilahti, J. Tuomilehto, Association of blood pressure and hypertension with the risk of Parkinson disease: the National FINRISK Study, Hypertension 57 (6) (2011) 1094e1100. [16] I.S. Bradvica, D. Janculjak, S. Butkovic-Soldo, A.M. Mrden, G. Kondza, R. Tadzic, Ultrasound evaluation of extracranial carotid artery lesions in Parkinsonian patients, Coll. Antropol. 34 (3) (2010) 1081e1085. [17] I. Rektor, D. Goldemund, K. Sheardova, I. Rektorova, Z. Michalkova, M. Dufek, Vascular pathology in patients with idiopathic Parkinson's disease, Park. Relat. Disord. 15 (1) (2009) 24e29. [18] M. Kawata, Y. Nemoto, M. Asahina, I. Moroo, M. Shinomiya, T. Yamada, Risk
[19]
[20]
[21]
[22] [23]
[24]
[25]
[26] [27] [28] [29]
[30]
[31]
[32]
[33]
factors for cerebral arteriosclerosis in Parkinson's disease, Park. Relat. Disord. 2 (2) (1996) 75e79. J.M. Lee, K.W. Park, W.K. Seo, M.H. Park, C. Han, I. Jo, S. Ahn Jo, Carotid intimamedia thickness in Parkinson's disease, Mov. Disord. 22 (16) (2007) 2446e2449. A. Elbaz, M. Ripert, B. Tavernier, B. Fevrier, M. Zureik, J. Gariepy, A. Alperovitch, C. Tzourio, Common carotid artery intima-media thickness, carotid plaques, and walking speed, Stroke 36 (10) (2005) 2198e2202. C. Ankle Brachial Index, F.G. Fowkes, G.D. Murray, I. Butcher, C.L. Heald, R.J. Lee, L.E. Chambless, A.R. Folsom, A.T. Hirsch, M. Dramaix, G. deBacker, J.C. Wautrecht, M. Kornitzer, A.B. Newman, M. Cushman, K. Sutton-Tyrrell, F.G. Fowkes, A.J. Lee, J.F. Price, R.B. d'Agostino, J.M. Murabito, P.E. Norman, K. Jamrozik, J.D. Curb, K.H. Masaki, B.L. Rodriguez, J.M. Dekker, L.M. Bouter, R.J. Heine, G. Nijpels, C.D. Stehouwer, L. Ferrucci, M.M. McDermott, H.E. Stoffers, J.D. Hooi, J.A. Knottnerus, M. Ogren, B. Hedblad, J.C. Witteman, M.M. Breteler, M.G. Hunink, A. Hofman, M.H. Criqui, R.D. Langer, A. Fronek, W.R. Hiatt, R. Hamman, H.E. Resnick, J. Guralnik, M.M. McDermott, Ankle brachial index combined with framingham risk score to predict cardiovascular events and mortality: a meta-analysis, Jama 300 (2) (2008) 197e208. V. Dias, E. Junn, M.M. Mouradian, The role of oxidative stress in Parkinson's disease, J. Park. Dis. 3 (4) (2013) 461e491. L. Loffredo, A. Marcoccia, P. Pignatelli, P. Andreozzi, M.C. Borgia, R. Cangemi, F. Chiarotti, F. Violi, Oxidative-stress-mediated arterial dysfunction in patients with peripheral arterial disease, Eur. Heart J. 28 (5) (2007) 608e612. V. Kotagal, R.L. Albin, M. Müller, R.A. Koeppe, K.A. Frey, N.I. Bohnen, Modifiable cardiovascular risk factors and axial motor impairments in Parkinson disease, Neurology 82 (17) (2014) 1514e1520. M. Asahina, E. Vichayanrat, D.A. Low, V. Iodice, C.J. Mathias, Autonomic dysfunction in parkinsonian disorders: assessment and pathophysiology, J. Neurol. Neurosurg. Psychiatry 84 (6) (2013) 674e680. D. Gupta, A. Kuruvilla, Vascular parkinsonism: what makes it different? Postgrad. Med. J. 87 (1034) (2011) 829e836. B. Thanvi, N. Lo, T. Robinson, Vascular parkinsonismean important cause of parkinsonism in older people, Age ageing 34 (2) (2005) 114e119. M. Critchley, Arteriosclerotic parkinsonism, Brain 52 (1929) 23e83. H.M. van der Holst, I.W.M. van Uden, A.M. Tuladhar, K.F. de Laat, A.G.W. van Norden, D.G. Norris, E.J. van Dijk, R.A.J. Esselink, B. Platel, F.-E. de Leeuw, Cerebral small vessel disease and incident parkinsonism: the RUN DMC study, Neurology 85 (18) (2015) 1569e1577. C. Reitz, C. Trenkwalder, K. Kretzschmar, A. Roesler, V. Eckardstein, K. Berger, Relation of cerebral small-vessel disease and brain atrophy to mild Parkinsonism in the elderly, Mov. Disord. 21 (11) (2006) 1914e1919. J. Hatate, K. Miwa, M. Matsumoto, T. Sasaki, Y. Yagita, M. Sakaguchi, K. Kitagawa, H. Mochizuki, Association between cerebral small vessel diseases and mild parkinsonian signs in the elderly with vascular risk factors, Park. Relat. Disord. 26 (2016) 29e34. R. Kromer, C. Buhmann, U. Hidding, M. Keserü, D. Keserü, A. Hassenstein, B. Stemplewitz, Evaluation of retinal vessel morphology in patients with Parkinson's disease using optical coherence tomography, PLoS One 11 (8) (2016) e0161136. A.H. Rajput, E.F. Rajput, Octogenarian parkinsonism - clinicopathological observations, Park. Relat. Disord. 37 (2017) 50e57.
Please cite this article in press as: V. Vlasov, et al., Subclinical vascular disease and the risk of parkinsonism: The Rotterdam Study, Parkinsonism and Related Disorders (2017), http://dx.doi.org/10.1016/j.parkreldis.2017.06.022
Contents lists available at ScienceDirect
Parkinsonism and Related Disorders journal homepage: www.elsevier.com/locate/parkreldis
Subclinical vascular disease and the risk of parkinsonism: The Rotterdam Study Vanja Vlasov a, Sirwan K.L. Darweesh a, b, c, Bruno H. Stricker a, d, Oscar H. Franco a, M.Kamran Ikram a, b, Maryam Kavousi a, Daniel Bos a, c, e, Caroline C.W. Klaver a, f, M.Arfan Ikram a, b, e, * a
Department of Epidemiology, Erasmus MC University Medical Center Rotterdam, Rotterdam, The Netherlands Department of Neurology, Erasmus MC University Medical Center Rotterdam, Rotterdam, The Netherlands Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, MA, USA d Inspectorate for Health Care, 2595 AN, The Hague, The Netherlands e Department of Radiology, Erasmus MC University Medical Center Rotterdam, Rotterdam, The Netherlands f Department of Ophthalmology, Erasmus MC University Medical Center Rotterdam, Rotterdam, The Netherlands b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 10 February 2017 Received in revised form 3 June 2017 Accepted 27 June 2017
Background: Parkinsonism is a common neurodegenerative syndrome in middle-aged and elderly persons. The etiology is multifactorial with a possible vascular contribution, but this has not been comprehensively studied. Objective: To determine whether selected markers of subclinical vascular pathology are associated with the risk of all-cause parkinsonism in the general population. Methods: We assessed a range of markers of subclinical vascular pathology (ankle-brachial index, carotid plaques and intima media thickness, retinal arteriolar and venular calibers) in 6199 persons from the population-based Rotterdam Study, who were free of parkinsonism and dementia at baseline. We followed these persons up till onset of parkinsonism, dementia, and death for 89,387 person-years until January 1, 2013. Hazard ratios (HRs) for all-cause parkinsonism and separately for Parkinson disease (PD) versus non-PD were estimated from competing risk regression models adjusting for potential confounders. Results: During follow-up, we identified 211 cases of parkinsonism (110 had PD). None of the five markers of subclinical pathology was associated with all-cause parkinsonism. Only low ankle-brachial index was associated with a higher risk of non-PD parkinsonism (HR ¼ 0.79, 95%CI: 0.68e0.92), but not with the risk of PD. Conclusion: We did not find a consistent pattern of associations between systemic vascular pathology markers with parkinsonism, suggesting that the potential involvement of vascular pathology is not prominent or needs further evaluation in studies with an even larger sample size. © 2017 Elsevier Ltd. All rights reserved.
Keywords: Parkinsonism Vascular pathology Risk factors Etiology
1. Introduction Parkinsonism, with its most frequent subtype Parkinson disease (PD), is a common chronic neurodegenerative syndrome in the elderly [1,2]. Although many genetic and environmental factors have been implicated in the multi-factorial etiology of parkinsonism [1], there is still much uncertainty about the exact
* Corresponding author. Department of Epidemiology, Erasmus MC University Medical Center, Dr. Molewaterplein 50, 3015 GE, Rotterdam, The Netherlands. E-mail address: [email protected] (M.Arfan Ikram).
mechanisms underlying neuronal cell loss in patients with Parkinson disease and non-PD parkinsonism. There is a scarcity of knowledge on early markers and potential neuroprotective therapies in early stages of parkinsonism, and patients are typically already in an advanced pathological disease stage when the diagnosis is made [1]. Therefore, there is a need for novel insight into factors associated with all-cause parkinsonism. Markers of subclinical vascular pathology are strongly implicated in the etiology of the two common neurological syndromes, stroke and dementia [3]. Furthermore, a high prevalence of lacunar infarcts have been reported in the cerebral white matter and basal
http://dx.doi.org/10.1016/j.parkreldis.2017.06.022 1353-8020/© 2017 Elsevier Ltd. All rights reserved.
Please cite this article in press as: V. Vlasov, et al., Subclinical vascular disease and the risk of parkinsonism: The Rotterdam Study, Parkinsonism and Related Disorders (2017), http://dx.doi.org/10.1016/j.parkreldis.2017.06.022
2
V. Vlasov et al. / Parkinsonism and Related Disorders xxx (2017) 1e6
ganglia of patients with parkinsonism [4], which indicates a potential vascular etiology. During the course of Alzheimer's disease 25% of patients develop parkinsonism [5], whereas approximately 30% of patients with PD are eventually diagnosed with dementia [6]. However, in spite of an overlap in clinical and pathological findings between primary parkinsonism (such as PD) and dementia or stroke, the role of vascular pathology in the etiology of parkinsonism syndromes remains unclear. Vascular pathology can be non-invasively measured both in large systemic vessels as well as in microvasculature locally [7]. Commonly used, non-invasive measurements of vascular pathology of the large vessels include ultrasound measures of carotid arteries (intima media thickness and plaques) and ankle-brachial index (ABI) [8,9]. Furthermore, structural changes in the retinal vasculature have long been recognized as an important marker of microvascular pathology [10]. These non-invasive measures provide an insight into the condition of the large and small blood vessels. We hypothesized that subclinical vascular pathology is associated with an increased risk of parkinsonism in the general population. We used noninvasive measurements of subclinical vascular disease to test our hypothesis in a large, population-based cohort study. 2. Methods 2.1. Setting and study population This study was conducted as part of the Rotterdam Study, a community-based prospective cohort study in elderly adults [11]. There were only two selection criteria: persons had to live in the well-defined Ommoord district of Rotterdam in 1990, and they had to be aged 55 years or older. Out of 10,215 eligible individuals, 7983 agreed to participate and provided informed consent. All participants were interviewed and underwent extensive physical examination at the research center, including screening for parkinsonian signs and markers of vascular pathology at baseline (1990e1993) and at four follow-up rounds (1993e1995, 1997e1999, 2002e2004, and 2009e2011). For the current study we excluded participants with prevalent parkinsonism (n ¼ 126), incomplete neurological screening (n ¼ 1401) at baseline, and participants with missing information regarding two or more vascular pathology markers (n ¼ 257), leaving 6199 subjects at risk to develop any kind of parkinsonism during the follow-up period. Similarly with other population-based studies, the Rotterdam Study overall age- and sex-adjusted incidence rate of Parkinson Disease is 1.5 per 1000 person-years [2], while age-specific incidence rates are somewhat higher than in community based studies [12]. The Rotterdam Study has been approved by the medical ethics committee according to the Population Screening Act: Rotterdam Study, executed by the Ministry of Health, Welfare and Sports of the Netherlands. All participants provided written informed consent. 2.2. Assessment of subclinical vascular disease Recently, CT and MRI assessments were implemented in the Rotterdam Study; however, we lacked sufficient follow-up for incident parkinsonism to incorporate these measurements in our present analyses. Subclinical vascular pathology was assessed with four different measures: ankle-brachial index, carotid intimamedia thickness and the presence of plaques in the carotid arteries assessed using ultrasonography and retinal vessels diameters.
2.3. Ankle-brachial index We assessed ankle-brachial index (ABI) by measuring systolic blood pressure (SBP) in all extremities. The SBP was measured twice in both arms and at both ankles (posterior tibial artery) in each participant in the supine position with an 8-MHx continuouswave Doppler device (Huntleigh D500, Hunteleigh Technology) which was attached to a standard random-zero sphygmomanometer [8]. To obtain the ABI we calculated the ratio of the systolic blood pressure at the ankle and the systolic blood pressure at the arm per side. ABI values higher than 1.50 were not included in the study, because that values are considered irrationally high due to calcification of the blood vessels and their inability to properly compress. We used the lowest value of the two sides for the analyses [8].
2.4. Carotid number of plaques and intima media thickness Using B-mode ultrasonography we evaluated both carotid arteries for the presence of plaques and determined the carotid intima media thickness (cIMT) [9]. The presence of a plaque was defined as a focal widening relative to bordering segments, with protrusion into the lumen consisting of either echolucent or echogenic areas. We assessed the presence of plaques at both sides in the common carotid artery, carotid bifurcation and proximal part of the internal carotid artery. The total amount of plaques was reflected in a weighted plaque score that ranged from 0 to 6 [13]. The average of the maximum IMT was registered at both the near- and far wall of the common carotid artery for assessing the cIMT.
2.5. Retinal vessel measurements At the baseline ophthalmic examination, simultaneous stereoscopic fundus color transparencies were taken centered on the optic disk (20 field, Topcon Optical Company, Tokyo, Japan) after pharmacologic mydriasis and were digitized with a high-resolution scanner (Nikon LS-4000, Nikon Corporation, Japan) [14]. Evaluated measures for arteriolar and venular calibers were based on improved Parr-Hubbard formulas and were corrected for magnification changes due to refractive errors of the eye [14]. Since each eye has a different magnification in case of refractive changes, summary vessel measures were additionally adjusted for the refraction and corneal curvature with Littmann's formula to estimate absolute measures [14].
2.6. Other measurements We assessed cardiovascular risk factors by interview, physical examination and blood sampling [11]. Blood pressure was calculated as the mean of two consecutive measurements with a random-zero sphygmomanometer at the right brachial artery while the patient was in a sitting position. Serum total cholesterol was determined by an automated enzymatic procedure in a non-fasting blood sample. Serum high density lipoprotein (HDL) cholesterol was measured after precipitation of the non-HDL fraction with phosphotungstate-magnesium. Height and weight were measured and body mass index (kg/m2) was calculated. Non fasting serum glucose levels were measured using the glucose hexokinase method. Smoking habits of participants were categorized as “ever” or “never”. Uric acid was determined with a Kone Diagnostica reagent kit and a Kone auto-analyzer from blood samples of participants.
Please cite this article in press as: V. Vlasov, et al., Subclinical vascular disease and the risk of parkinsonism: The Rotterdam Study, Parkinsonism and Related Disorders (2017), http://dx.doi.org/10.1016/j.parkreldis.2017.06.022
V. Vlasov et al. / Parkinsonism and Related Disorders xxx (2017) 1e6
2.7. Ascertainment of parkinsonism and parkinson disease A detailed description of parkinsonism and Parkinson disease assessment methods has previously been published [2] and is summarized in Supplement 1. In short, we used four overlapping modalities to screen for potential parkinsonism during follow-up: in-person screening (on average every 4 years), in-person interviews, use of antiparkinsonian medication, and alerts from continuous monitoring of clinical records. Of all persons who screened positive in any of these methods, complete medical records (including letters from medical records of specialists and general practitioners) were studied and case reports were drawn up covering all potentially relevant information to establish presence and cause of parkinsonism. These case reports were evaluated by a panel led by an experienced neurologist. Parkinson disease was only diagnosed after exclusion of secondary causes, and medical records of all incident parkinsonism cases (both PD and non-PD) continued to be scrutinized until the end of the study period for new information that could lead to a revision of the diagnosis. Given the substantial overlap between the four detection methods we considered persons who were not screened in-person during one of the follow-up rounds still at risk for parkinsonism and Parkinson disease. Person-time at risk of incident parkinsonism ended at onset of parkinsonism, death, or January 1, 2013. Person-time at risk of incident Parkinson disease ended at onset of parkinsonism, incident dementia (date of clinical diagnosis), stroke, death, or January 1, 2013. 2.8. Statistical analysis Baseline characteristics are presented as frequencies with percentages for categorical variables or means (standard deviation [SD]) for continuous variables. Potential differences in baseline characteristics between included and excluded individuals were explored using independent t-tests for continuous variables and chi-square tests for categorical variables (Supplement 2). The sample size and power calculations were calculated a priori and presented in Supplement 3. The association of measures of subclinical vascular disease with risk of parkinsonism was investigated using competing risk regression models. We analyzed each risk factor separately. All continuous vascular pathology markers were transformed into zscores to facilitate direct comparison of effect sizes. We performed similar analyses for the risk of Parkinson disease and non-PD parkinsonism. For non-PD parkinsonism, we considered PD and death to be competing events; for PD, we considered non-PD parkinsonism, death, stroke and dementia as a competing events. All analyses were adjusted for age and sex (model 1). Additionally, we adjusted our analyses for systolic blood pressure, diastolic blood pressure, body mass index, total cholesterol level, HDL cholesterol level, glucose level, smoking status and serum uric acid (model 2). In all analyses involving retinal vessel diameters we adjusted these for each other (venular diameters for arterial diameters and vice versa). To explore possible effect modification in our analysis we separately added interaction terms of sex and age with markers of subclinical vascular pathology (i.e. sex*ABI; age*ABI) and conducted stratified analyses on sex and on age (using mean age as the threshold). To investigate whether association of ABI and non-PD parkinsonism were driven by drug-induced parkinsonism, we repeated analyses for association after excluding 20 medicationinduced cases. In order to assess subclinical vascular disease measures as prodromal markers instead of risk factors, we repeated main analyses after making cut offs in a follow-up time on two years, five years and ten years (Supplement 4). We conducted sensitivity analysis on
3
estimates of the association between subclinical vascular markers and all-cause parkinsonism after censoring cases of unspecified parkinsonism. Moreover, we stratified Parkinson disease patients into groups with and without subclinical vascular pathology, and compare them according to the age of onset of the disease, UPDRSIII score, treatment and a mortality rate from the date of onset of parkinsonism until the date of censoring. We had complete data on 95% of independent variables. Missing values were imputed by multiple imputation (n ¼ 5), based on gender, age and all others covariables. IBM SPSS Statistics version 21.0 (IBM Corp, Armonk, NY, USA) and R version 3.2.5 (R Foundation for Statistical Computing, Vienna, Austria) were used for statistical analyses. 3. Results Baseline characteristics of the study population are presented in Table 1. The mean age at baseline of the eligible 6199 people was 68.5 years and 58.7% of the participants was female. A comparison between included and excluded individuals demonstrated that excluded individuals were older, had a significantly higher body mass index, higher ABI, larger cIMT, and higher serum glucose level (Supplement 2. Table 1.). During a total follow-up of 89,387 person-years, [mean (SD), 14.42 (6.5)], we identified 211 patients with parkinsonism, of whom 110 patients were diagnosed with Parkinson disease. An overview of diagnoses is provided in Table 2. Table 3 shows the association between markers of subclinical vascular disease and the risk of all-cause parkinsonism, Parkinson disease and non-PD parkinsonism. Overall, no measure of subclinical vascular disease was associated with the risk of all-cause parkinsonism. Correspondingly, none of the vascular markers in our study were associated with risk of Parkinson disease. We only found a decreased risk for non-PD parkinsonism for participants with higher ABI [HR: 0.79 (95% CI: 0.68e0.92)]. After excluding drug-induced cases, associations of ABI and the non-PD parkinsonism remained almost unchanged [HR: 0.77 (95% CI: 0.63e0.93)]. Furthermore, carotid plaques, cIMT and diameter of retinal vessels were not associated with non-PD parkinsonism. Additional adjustment for cardiovascular risk factors did not substantially alter any of these results (Table 3, Model 2). We found no evidence for statistical interaction by sex or age (all p-interaction>0.73). Additional analyses of subclinical vascular pathology markers as predictors of all-cause parkinsonism over restricted follow-up
Table 1 Baseline characteristics of the study population. Characteristic
Participants (n ¼ 6199)
Age at baseline, years Women, % Smoking, % Body mass index, kg/m2 Total cholesterol, mmol/L HDL cholesterol, mmol/L Systolic blood pressure, mm Hg Diastolic blood pressure, mm Hg Serum glucose, mmol/L Serum uric acid, mmol/L Ankle-brachial index Weighted number of plaques, number Carotid intima media thickness, mm Retinal arteriolar diameter, mm Retinal venular diameter, mm
68.5 ± 8.5 58.7 64.5 26.3 ± 3.7 6.6 ± 1.2 1.3 ± 0.3 139.2 ± 22.2 73.8 ± 11.4 6.8 ± 2.5 321 ± 80.7 1.0 ± 0.2 1.5 ± 1.6 1.0 ± 0.2 146.7 ± 14.3 221.7 ± 20.9
HDL, high-density lipoprotein. Values are mean ± standard deviation for continuous variables, percentage for dichotomous variables.
Please cite this article in press as: V. Vlasov, et al., Subclinical vascular disease and the risk of parkinsonism: The Rotterdam Study, Parkinsonism and Related Disorders (2017), http://dx.doi.org/10.1016/j.parkreldis.2017.06.022
4
V. Vlasov et al. / Parkinsonism and Related Disorders xxx (2017) 1e6
characteristics between the two groups. We cannot exclude that selection bias led to slight underestimations of associations between markers of subclinical vascular pathology and parkinsonism. Moreover, we note that we excluded individuals who were younger than 55 years at baseline, and that this selection could have introduced bias if the association between subclinical vascular disease and incident PD is substantially different in younger individuals. Second, it seems unlikely that unmeasured confounding could account for the null results, but it cannot be completely excluded in observational studies. The role of potential confounders that were assessed was examined in multivariate models, and these were found to have little impact on effect estimates. Third, we used only extracranial measures of subclinical vascular pathology and not the most sensitive markers of subclinical vascular pathology. In the future, community-based studies with adequate follow-up for incident parkinsonism should focus on measures of intracranial vascular pathology imaging, including novel brain-imaging techniques (e.g. calcification imaging on CT, arterial spin labeling on fMRI). Moreover, the vast majority of an unspecified parkinsonism in our study did not have neuroimaging information, those patients could not be classified as PD or other parkinsonism subtypes. Finally, some misclassification of Parkinson Disease cases as other subtypes of parkinsonism might have occurred, since we had no pathological confirmation of the diagnoses. This prospective population-based study has several essential and unique strengths. First, all participants were included and followed up using the same methodology, irrespective of their risk of parkinsonism. Second, the large sample size, long follow-up duration and the carefully collected prospective data on exposure variables as well as on potential confounders minimizes misclassification of baseline vascular pathology profiles of study participants. Furthermore, we actively screened participants for parkinsonism at baseline and only included persons without parkinsonism. Since all assessments of vascular pathology markers took place at baseline, our findings unequivocally reflect the prediagnostic phase of parkinsonism and Parkinson disease. Finally, our near-complete follow-up for dementia and death allowed us to account for events that possibly interfere with the observation of the parkinsonism and Parkinson disease. Previous population-based study has focused mainly on the association between hypertension and the risk of Parkinson disease [15]. However, there is a little prospective data on the role of subclinical vascular disease in the etiology of parkinsonism in the
Table 2 Overview of incident clinical diagnoses of parkinsonism. All-cause PS patients (n ¼ 211) Parkinson disease Unspecified parkinsonism Drug-induced parkinsonism Vascular parkinsonism Multiple system atrophy Lewy body dementia Secondary to dementia Secondary to a tumor Progressive supranuclear palsy
110 (52.1%) 61 (28.9%) 20 (9.5%) 8 (3.8%) 4 (1.9%) 1 (0.5%) 4 (1.9%) 2 (0.9%) 1 (0.5%)
These are clinical diagnoses; no histologic confirmation was obtained. n, number of cases during follow-up.
periods are presented in the supplementary 4, Table 3. In analyses that were restricted on 2-years of follow up, hazard ratios were slightly higher than in the overall analyses. Results from other time-restricted analyses did not differ from the overall analyses. The Supplementary Table 4 shows the results of main analyses after censoring 61 cases with unspecified parkinsonism. No differences were found between the main results and the results after censoring unspecified parkinsonism cases. Parkinson disease patients, who were in the top risk-increasing quartile of at least one subclinical vascular pathology marker, had later onset of a parkinsonism comparing with Parkinson disease patients without vascular pathology. There was no difference regarding to mortality rate, UPDRS-III score and medication use between this two groups.
4. Discussion We did not find an association between markers of subclinical vascular disease and the risk of all-cause parkinsonism or Parkinson disease. For non-PD parkinsonism, we only observed a significant association with lower ABI, but not with any other marker of subclinical vascular disease. Several limitations of this study need to be discussed. First, we excluded participants who were not screened for parkinsonism at baseline and persons with missing data on more than two markers of subclinical vascular pathology. Participants with a missing score were older, on average with lower ABI and with higher cIMT compared to those who were eligible for the study (Supplement 2, Table 1.). No significant differences were found in other measured
Table 3 Associations of markers of subclinical vascular pathology with the risk of parkinsonism. Parkinsonism n/N ¼ 211/6199
Parkinson Disease n/N ¼ 110/6199
Non PD parkinsonism n/N ¼ 101/6199
HR (95% CI)
HR (95% CI)
HR (95% CI)
1.14 0.84 0.86 0.98 0.96
(0.94e1.39) (0.68e1.05) (0.68e1.10) (0.77e1.25) (0.76e1.23)
0.79 (0.68e0.92)* 0.97 (0.80e1.18) 0.97 (0.77e1.22) 0.89 (0.68e1.17) 0.89 (0.67e1.19)
1.08 0.90 0.91 1.04 1.01
(0.87e1.33) (0.72e1.14) (0.69e1.19) (0.82e1.31) (0.78e1.27)
0.76 (0.65e0.90) 0.99 (0.82e1.20) 0.98 (0.78e1.24) 0.96 (0.73e1.26) 0.86 (0.64e1.16)
Per 1-SD increase in:
Model 1
Ankle brachial index Weighted number of plaques Carotid intima media thickness Retinal arteriolar diameter Retinal venular diameter
0.93 0.90 0.91 0.94 0.92
(0.82e1.06) (0.78e1.05) (0.77e1.08) (0.78e1.13) (0.77e1.12)
Model 2 Ankle brachial index Weighted number of plaques Carotid intima media thickness Retinal arteriolar diameter Retinal venular diameter
0.88 0.95 0.94 1.01 0.93
(0.77e1.01) (0.81e1.10) (0.79e1.13) (0.84e1.20) (0.77e1.12)
n, number of cases; N, number of persons at risk; HR, hazard ratio; CI, confidence interval; SD, standard deviation. Model 1: Adjusted for age and sex. Model 2: Adjusted for age, sex, systolic blood pressure, diastolic blood pressure, smoking, body mass index, total cholesterol, HDL cholesterol, glucose, uric acid. *Significant (p < 0.05) data in bold.
Please cite this article in press as: V. Vlasov, et al., Subclinical vascular disease and the risk of parkinsonism: The Rotterdam Study, Parkinsonism and Related Disorders (2017), http://dx.doi.org/10.1016/j.parkreldis.2017.06.022
V. Vlasov et al. / Parkinsonism and Related Disorders xxx (2017) 1e6
general population. Moreover, previous observations were focused on whether vascular pathology contribute to the disease severity in already diagnosed PD patients [16,17] or whether vascular disease co-occurred with Parkinson disease [18,19]. The main novelty of the current study is that we studied markers of several subclinical vascular pathology as risk factors of incident parkinsonism. Remarkably, none of these measures were associated with all-cause parkinsonism and Parkinson disease. Carotid pathology as shown on ultrasound was associated with motor and the cognitive impairment in the PD patients in previous studies [16,17]. Additionally, carotid plaques and higher cIMT have been associated with gait and balance decline over time [20]. Also, patients with vascular parkinsonism have a higher risk for embolic cerebrovascular events [16]. However, another study showed that patients with Parkinson disease had less severe carotid lesions with mostly stable calcified plaques compared to the general population [19]. In this study, we investigated the cIMT and number of plaques as potential risk factors of onset of parkinsonism and Parkinson disease. We observed that ultrasound measurements of carotid artery pathology were not related to a higher risk of all-cause parkinsonism, Parkinson disease or non-PD parkinsonism. A commonly used measurement of vascular pathology in symptomatic and asymptomatic individuals with cardiovascular disease is ABI [8,21]. Ankle brachial index is inversely related to the severity of peripheral arterial disease (PAD), which mean that persons with lower ABI have more expressed pathology of lower extremity arteries [8]. We found that ABI was not related to a higher risk of Parkinson disease but instead with the risk of non-PD parkinsonism. There are several possible explanations for the link between peripheral arterial disease (PAD) and non-PD parkinsonism, such as oxidative stress as same pathophysiological mechanism [22,23] and that vascular pathology could lead to neurodegeneration outside of the nigrostriatal area in non-PD parkinsonism patients [24]. Autonomic dysfunction can be one of the first symptoms in non-PD parkinsonism patients [25]. Moreover, the autonomic nervous system (ANS) has blood supply from peripheral vessels and PAD might lead to neuronal cell loss of ANS. Vascular parkinsonism (VP) can occur after a stroke and studies suggest that atherosclerotic changes have an influence on this type of parkinsonism [26,27]. More than eighty years ago arteriosclerotic parkinsonism was already recognized as a distinct clinical entity by Critchley [28]. While percentage of VP cases among non-PD parkinsonism in our study is low (7.9%), we cannot exclude misclassification of VP as unspecified parkinsonism regarding its much higher percentage in the same substrate (60.4%). Therefore, the association of unspecified parkinsonism with peripheral vascular pathology might warrant further investigation. However, these results should be interpreted with caution because of the lack of associations of other vascular pathology markers and the risk of non-PD parkinsonism. Any impact of multiple testing cannot be ruled out. The diameter of retinal vessels is mainly associated with cerebral small vessel disease (CSVD) [10]. In our study there was no relationship between arteriolar diameters and larger venular diameters with all-cause parkinsonism. This is contrary to two crosssectional studies and one study in patients with CSVD, which revealed that measures of CSVD are associated with the incidence of parkinsonian signs [29e31]. Those studies were using more sensitive measures of CSVD, such as white matter volume, number of lacunes, and cerebral microbleeds. However, their follow-up time was nonexistent or short, and their sample size was smaller than in our study. Similar to our results, one case-control study showed that retinal vessel diameter did not differ between PD patients and healthy controls [32]. Additionally, recent clinicopathological study observed no difference between the incidence of
5
small vessels disease pathology and stroke in PD cases compared with the general population [33]. Different methodology and different types of parameters for CSVD can explain the divergent results from studies on CSVD and parkinsonism risk. In our cohort of middle-aged and elderly adults, potential risk factors for late-onset of parkinsonism were investigated. Still, we cannot exclude the possibility that the subclinical vascular pathology might be associated with young-onset of parkinsonism. In conclusion, we did not find a consistent pattern of associations between measures of subclinical vascular pathology and allcause parkinsonism, thus our report suggest that systemic vascular pathology does not play a large role in the etiology of parkinsonism in the general population. Still, prospective studies that incorporate novel brain-imaging and have adequate sample size and follow-up may detect associations of intracranial vascular pathology with incident parkinsonism. Author contributions Study concept and design: Vanja Vlasov, Sirwan Darweesh, Daniel Bos. Acquisition of data: Arfan Ikram, Sirwan Darweesh, Daniel Bos, Vanja Vlasov. Statistical analysis and interpretation of data: Vanja Vlasov. Drafting the manuscript: Vanja Vlasov. Critical revision of the manuscript for important intellectual content: Arfan Ikram, Kamran Ikram, Bruno Sticker, Oscar Franco, Maryam Kavousi and Caroline Klaver. Study supervision: Sirwan Darweesh, Daniel Bos, Arfan Ikram. Sources of funding This study received support from Stichting ParkinsonFonds. The Rotterdam Study is further supported by the Erasmus MC University Medical Center and Erasmus University Rotterdam, the Netherlands Organization for Scientific Research (NWO), the Netherlands Organization for Health Research and Development (ZonMW), the Research Institute for Diseases in the Elderly (RIDE), the Ministry of Education, Culture and Science, the Ministry of Health, Welfare and Sport, The European Commission (DGXII), the Netherlands Genomics Initiative (NGI) and the Municipality of Rotterdam. The contribution of the inhabitants, general practitioners and pharmacists of the Ommoord district to the Rotterdam Study are gratefully acknowledged. Disclosure statement All authors have approved the final article and report no conflicts of interest. The paper or any part of it has not been published previously and is not under consideration for publication elsewhere. None of the authors report disclosures. Conflicts of interest None. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.parkreldis.2017.06.022. References [1] L.M. de Lau, M.M. Breteler, Epidemiology of Parkinson's disease, Lancet Neurol. 5 (6) (2006) 525e535.
Please cite this article in press as: V. Vlasov, et al., Subclinical vascular disease and the risk of parkinsonism: The Rotterdam Study, Parkinsonism and Related Disorders (2017), http://dx.doi.org/10.1016/j.parkreldis.2017.06.022
6
V. Vlasov et al. / Parkinsonism and Related Disorders xxx (2017) 1e6
[2] S.K. Darweesh, P.J. Koudstaal, B.H. Stricker, A. Hofman, M.A. Ikram, Trends in the incidence of Parkinson disease in the general population: the Rotterdam study, Am. J. Epidemiol. 183 (11) (2016) 1018e1026. [3] H. Gardener, C.B. Wright, T. Rundek, R.L. Sacco, Brain health and shared risk factors for dementia and stroke, Nat. Rev. Neurol. 11 (11) (2015) 651e657. [4] E.E. Smith, J.A. Schneider, J.M. Wardlaw, S.M. Greenberg, Cerebral microinfarcts: the invisible lesions, Lancet Neurol. 11 (3) (2012) 272e282. [5] F. Portet, N. Scarmeas, S. Cosentino, E.P. Helzner, Y. Stern, Extrapyramidal signs before and after the diagnosis of incident Alzheimer disease in a prospective population study, Arch. Neurol. 66 (9) (2009) 1120e1126. [6] D. Aarsland, K. Andersen, J.P. Larsen, A. Lolk, P. Kragh-Sorensen, Prevalence and characteristics of dementia in Parkinson disease: an 8-year prospective study, Arch. Neurol. 60 (3) (2003) 387e392. [7] S. Laurent, M. Briet, P. Boutouyrie, Large and small artery cross-talk and recent morbidity-mortality trials in hypertension, Hypertension 54 (2) (2009) 388e392. [8] F.G. Fowkes, The measurement of atherosclerotic peripheral arterial disease in epidemiological surveys, Int. J. Epidemiol. 17 (2) (1988) 248e254. [9] M.L. Bots, A. Hofman, P.T. De Jong, D.E. Grobbee, Common carotid intimamedia thickness as an indicator of atherosclerosis at other sites of the carotid artery. The Rotterdam Study, Ann. Epidemiol. 6 (2) (1996) 147e153. [10] M.K. Ikram, F.J. De Jong, E.J. Van Dijk, N.D. Prins, A. Hofman, M.M. Breteler, P.T. De Jong, Retinal vessel diameters and cerebral small vessel disease: the Rotterdam Scan Study, Brain 129 (Pt 1) (2006) 182e188. [11] A. Hofman, G.G.O. Brusselle, S.D. Murad, C.M. van Duijn, O.H. Franco, A. Goedegebure, M.A. Ikram, C.C.W. Klaver, T.E.C. Nijsten, R.P. Peeters, B.H.C. Stricker, H.W. Tiemeier, A.G. Uitterlinden, M.W. Vernooij, The Rotterdam Study: 2016 objectives and design update, Eur. J. Epidemiol. 30 (2015) 661e708. [12] L.M. de Lau, P.C. Giesbergen, M.C. de Rijk, A. Hofman, P.J. Koudstaal, M.M. Breteler, Incidence of parkinsonism and Parkinson disease in a general population: the Rotterdam Study, Neurology 63 (7) (2004) 1240e1244. [13] I.M. van der Meer, M.L. Bots, A. Hofman, A.I. del Sol, D.A. van der Kuip, J.C. Witteman, Predictive value of noninvasive measures of atherosclerosis for incident myocardial infarction: the Rotterdam Study, Circulation 109 (9) (2004) 1089e1094. [14] L.D. Hubbard, R.J. Brothers, W.N. King, L.X. Clegg, R. Klein, L.S. Cooper, A.R. Sharrett, M.D. Davis, J. Cai, Methods for evaluation of retinal microvascular abnormalities associated with hypertension/sclerosis in the Atherosclerosis Risk in Communities Study, Ophthalmology 106 (12) (1999) 2269e2280. [15] C. Qiu, G. Hu, M. Kivipelto, T. Laatikainen, R. Antikainen, L. Fratiglioni, P. Jousilahti, J. Tuomilehto, Association of blood pressure and hypertension with the risk of Parkinson disease: the National FINRISK Study, Hypertension 57 (6) (2011) 1094e1100. [16] I.S. Bradvica, D. Janculjak, S. Butkovic-Soldo, A.M. Mrden, G. Kondza, R. Tadzic, Ultrasound evaluation of extracranial carotid artery lesions in Parkinsonian patients, Coll. Antropol. 34 (3) (2010) 1081e1085. [17] I. Rektor, D. Goldemund, K. Sheardova, I. Rektorova, Z. Michalkova, M. Dufek, Vascular pathology in patients with idiopathic Parkinson's disease, Park. Relat. Disord. 15 (1) (2009) 24e29. [18] M. Kawata, Y. Nemoto, M. Asahina, I. Moroo, M. Shinomiya, T. Yamada, Risk
[19]
[20]
[21]
[22] [23]
[24]
[25]
[26] [27] [28] [29]
[30]
[31]
[32]
[33]
factors for cerebral arteriosclerosis in Parkinson's disease, Park. Relat. Disord. 2 (2) (1996) 75e79. J.M. Lee, K.W. Park, W.K. Seo, M.H. Park, C. Han, I. Jo, S. Ahn Jo, Carotid intimamedia thickness in Parkinson's disease, Mov. Disord. 22 (16) (2007) 2446e2449. A. Elbaz, M. Ripert, B. Tavernier, B. Fevrier, M. Zureik, J. Gariepy, A. Alperovitch, C. Tzourio, Common carotid artery intima-media thickness, carotid plaques, and walking speed, Stroke 36 (10) (2005) 2198e2202. C. Ankle Brachial Index, F.G. Fowkes, G.D. Murray, I. Butcher, C.L. Heald, R.J. Lee, L.E. Chambless, A.R. Folsom, A.T. Hirsch, M. Dramaix, G. deBacker, J.C. Wautrecht, M. Kornitzer, A.B. Newman, M. Cushman, K. Sutton-Tyrrell, F.G. Fowkes, A.J. Lee, J.F. Price, R.B. d'Agostino, J.M. Murabito, P.E. Norman, K. Jamrozik, J.D. Curb, K.H. Masaki, B.L. Rodriguez, J.M. Dekker, L.M. Bouter, R.J. Heine, G. Nijpels, C.D. Stehouwer, L. Ferrucci, M.M. McDermott, H.E. Stoffers, J.D. Hooi, J.A. Knottnerus, M. Ogren, B. Hedblad, J.C. Witteman, M.M. Breteler, M.G. Hunink, A. Hofman, M.H. Criqui, R.D. Langer, A. Fronek, W.R. Hiatt, R. Hamman, H.E. Resnick, J. Guralnik, M.M. McDermott, Ankle brachial index combined with framingham risk score to predict cardiovascular events and mortality: a meta-analysis, Jama 300 (2) (2008) 197e208. V. Dias, E. Junn, M.M. Mouradian, The role of oxidative stress in Parkinson's disease, J. Park. Dis. 3 (4) (2013) 461e491. L. Loffredo, A. Marcoccia, P. Pignatelli, P. Andreozzi, M.C. Borgia, R. Cangemi, F. Chiarotti, F. Violi, Oxidative-stress-mediated arterial dysfunction in patients with peripheral arterial disease, Eur. Heart J. 28 (5) (2007) 608e612. V. Kotagal, R.L. Albin, M. Müller, R.A. Koeppe, K.A. Frey, N.I. Bohnen, Modifiable cardiovascular risk factors and axial motor impairments in Parkinson disease, Neurology 82 (17) (2014) 1514e1520. M. Asahina, E. Vichayanrat, D.A. Low, V. Iodice, C.J. Mathias, Autonomic dysfunction in parkinsonian disorders: assessment and pathophysiology, J. Neurol. Neurosurg. Psychiatry 84 (6) (2013) 674e680. D. Gupta, A. Kuruvilla, Vascular parkinsonism: what makes it different? Postgrad. Med. J. 87 (1034) (2011) 829e836. B. Thanvi, N. Lo, T. Robinson, Vascular parkinsonismean important cause of parkinsonism in older people, Age ageing 34 (2) (2005) 114e119. M. Critchley, Arteriosclerotic parkinsonism, Brain 52 (1929) 23e83. H.M. van der Holst, I.W.M. van Uden, A.M. Tuladhar, K.F. de Laat, A.G.W. van Norden, D.G. Norris, E.J. van Dijk, R.A.J. Esselink, B. Platel, F.-E. de Leeuw, Cerebral small vessel disease and incident parkinsonism: the RUN DMC study, Neurology 85 (18) (2015) 1569e1577. C. Reitz, C. Trenkwalder, K. Kretzschmar, A. Roesler, V. Eckardstein, K. Berger, Relation of cerebral small-vessel disease and brain atrophy to mild Parkinsonism in the elderly, Mov. Disord. 21 (11) (2006) 1914e1919. J. Hatate, K. Miwa, M. Matsumoto, T. Sasaki, Y. Yagita, M. Sakaguchi, K. Kitagawa, H. Mochizuki, Association between cerebral small vessel diseases and mild parkinsonian signs in the elderly with vascular risk factors, Park. Relat. Disord. 26 (2016) 29e34. R. Kromer, C. Buhmann, U. Hidding, M. Keserü, D. Keserü, A. Hassenstein, B. Stemplewitz, Evaluation of retinal vessel morphology in patients with Parkinson's disease using optical coherence tomography, PLoS One 11 (8) (2016) e0161136. A.H. Rajput, E.F. Rajput, Octogenarian parkinsonism - clinicopathological observations, Park. Relat. Disord. 37 (2017) 50e57.
Please cite this article in press as: V. Vlasov, et al., Subclinical vascular disease and the risk of parkinsonism: The Rotterdam Study, Parkinsonism and Related Disorders (2017), http://dx.doi.org/10.1016/j.parkreldis.2017.06.022