Figures
Abstract
Background
C-reactive protein (CRP) is a biomarker of inflammation, and high levels of CRP correlate with vascular death. Chronic inflammation is considered to be involved in neurodegeneration, although there is no evidence linking it with the process of neurodegenerative diseases.
Objective
To determine the role of baseline CRP levels in the prognosis of patients with Parkinson disease (PD).
Methods
A cohort of 313 patients with a mean age of 69.1 and mean PD duration of 7.9 years was retrospectively followed for a mean observation time of 1,753 days. CRP was measured when patients were not diagnosed with any infections, and levels were repetitively measured to investigate a tendency of “regression to mean.” The primary outcome measure was a survival time from study enrollment to death.
Results
During the observation period 56 patients died. Baseline CRP was log-linearly associated with a risk of death in PD. Mean survival time was 3,149 (95% confidence interval; 3,009-3,289) days in patients with CRP ≤ 0.8mg/L (lower two thirds) and 2,620 (2,343-2,897) days in those with CRP > 0.8 mg/L (top third, p < 0.001, log-rank test). The adjusted hazard ratio (HR) per two-fold higher CRP concentration for all deaths was 1.29 (1.10-1.52), and after excluding PD-unrelated deaths, such as cancer or stroke, HR was 1.23 (1.01-1.49) (adjusted for age, sex, PD duration, modified Hohen-Yahr stages, MMSE scores, and serum albumin).
Conclusions
Baseline CRP concentrations were associated with the risk of death and predicted life prognosis of patients with PD. The associations were independent from PD duration, PD severity, cognitive function, ages, and nutritional conditions, suggesting the possibility that subclinical chronic inflammation is associated with a neurodegenerative process in PD.
Citation: Sawada H, Oeda T, Umemura A, Tomita S, Kohsaka M, Park K, et al. (2015) Baseline C-Reactive Protein Levels and Life Prognosis in Parkinson Disease. PLoS ONE 10(7): e0134118. https://doi.org/10.1371/journal.pone.0134118
Editor: Kenji Hashimoto, Chiba University Center for Forensic Mental Health, JAPAN
Received: April 30, 2015; Accepted: July 6, 2015; Published: July 28, 2015
Copyright: © 2015 Sawada et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: All data files are available from the Dryad database: http://dx.doi.org/10.5061/dryad.63vc5.
Funding: Supported by Grants-in-Aid from the National Hospital Organization in Japan. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The level of C-reactive protein (CRP) in the peripheral blood is used as a biomarker of systemic inflammation, and is associated with increased risk of coronary heart disease [1], vascular death [2], and cancer death [3]. Baseline CRP levels (except for acute inflammation) are stable, similar to those of blood pressure or serum cholesterol [4].
While chronic inflammation is thought to be associated with neurodegeneration [5, 6], its relation to the prognosis of patients with brain diseases remains uncertain. Previous studies have reported a causal relationship between systemic inflammation and functional changes in the brain, such as “sickness behavior” [5, 7], as well as microglia activation in affected brain areas in Alzheimer disease [1], together with the generation of CRP in these brain lesions [8].
Parkinson disease (PD) is a neurodegenerative disorder characterized pathologically by dopaminergic neuronal death and the presence of Lewy bodies. Although the exact cause of the disease remains to be identified, neuroinflammatory mechanisms may contribute to the neurodegenerative disease process [9]. Pathological studies have demonstrated microglial activation in autopsied PD brains [10, 11], and this finding has been confirmed in neuroimaging studies using PK-11159, a marker of microglia activation, especially in the early stages of PD [12]. CRP is synthesized in hepatocytes, and plasma CRP concentrations are transiently and significantly elevated during acute inflammation [13]. In contrast, CRP concentrations are stable during the non-inflammatory phase [4]. A molecular and cellular communication between peripheral inflammation and the brain have been proposed [14]. In this context, it is reasonable to assume that CRP levels at baseline, i.e., in non-inflammatory conditions, are stable, and that such levels influence long-term prognosis by modulating the neurodegenerative process.
Although functional prognosis in PD is typically evaluated using indexes related to various clinical features, such as motor function, motor complications, autonomic failure, and cognitive decline, these indexes could be subjective and biased. Life prognosis is a solid endpoint with a definite inter-rater reliability and is, therefore, a suitable endpoint in retrospective analysis. In this study, we focused on life prognosis instead of functional prognosis to investigate the relationship between CRP and PD prognosis. We hypothesized that baseline CRP levels are associated with the neurodegenerative process. To test this hypothesis, we conducted a retrospective cohort study and determined the association between baseline plasma CRP levels and PD life prognosis.
Methods
Study design
For the purpose of the study, a retrospective cohort of PD patients was followed to determine the relationship between baseline CRP concentrations and PD life prognosis. First, we compared survival time of patients according to sex (male vs. female), age (young vs. elderly), PD disease duration (short vs. long), PD severity (modified Hoehn-Yahr stage), nutritional condition (high serum albumin vs. low serum albumin), and baseline CRP concentrations (low vs. high) at study enrollment. Then, the association between CRP concentrations and life prognosis was statistically examined using a hazard ratio according to the Cox hazard proportional model.
The study was approved by the Bioethics Committee of National Hospital of Utano (approval no. 26–4). The Bioethics committee waived the need for informed consent due to the retrospective nature of the study and anonymity of the collected data.
Patients
Consecutive patients with PD, who were treated at the Department of Neurology, the National Regional Center for Neurological Disorders and National Hospital of Utano from March 2004 to November 2007, were enrolled, and the medical records were retrospectively reviewed. PD diagnosis was based on the United Kingdom Parkinson’s disease Brain Bank Diagnostic Criteria. The enrollment criteria included PD patients free of any infection and who underwent measurements of plasma CRP.
Chronological stability of baseline CRP
To investigate the tendency for “regression to mean,” plasma CRP was repeatedly measured when the patients were free of any infection. Patients were assigned into three groups according to plasma CRP levels upon study enrollment (low, middle, and high), and follow-up CRP concentrations were measured 4 times at most (21–180 days, 181–360 days, 361–720 days, and 721–1,080 days), when patients were free of any infections for 28 days prior to and after blood sampling. Follow-up CRP concentrations were separately investigated according to the bottom, middle, and top thirds of baseline CRP concentrations to investigate the stability of baseline CRP concentrations. The definition of “free of infection” included no use of antibiotics, no fever (body temperature > 37.5°C), and no findings of pneumonia upon chest X-ray.
Survival time analysis
In survival time analysis, the observation period represented the time from study enrollment to the endpoint (date of death or May 16, 2014). Patients were censored when they were transferred to another hospital, and it was uncertain whether they were alive or dead. The study enrollment was the date of the first blood sampling for CRP measurements, which was performed for a routine blood test. Because of the possible association between CRP level baseline features, such as age, sex, PD duration and severity, cognitive function (Mini-Mental State Examination (MMSE), and nutritional condition (serum albumin level), we investigated the association using two methods, multi-variable analysis and stratifying analysis, according to clinical factors. Concerning the former, the association of baseline CRP levels was estimated using the hazard ratio (HR) according to the Cox proportional hazard model, which was adjusted for age, sex, mH-Y, PD duration, MMSE score, and serum albumin concentrations. In PD, the most common cause of death is pneumonia [15], and suffocation, fracture due to falls, dehydration, and unexpected sudden death are often seen in PD [16]; these causes can be considered to be related to PD. However, in contrast, vascular deaths or cancer deaths are not related to PD. Nevertheless, we examined the associations for all deaths. In addition, the associations with CRP and death, except for PD-unrelated events, were examined, and deaths from PD-unrelated events were censored or labeled “alternative outcomes.”
The following factors were collected for prediction variables; age, sex, PD duration, modified Hoehn-Yahr stages (mH-Y), MMSE scores, serum albumin, and plasma CRP concentrations upon study enrollment. In patients with motor fluctuations, mH-Y was evaluated in “ON period.” Although age and PD duration increased, and mH-Y and MMSE worsened during the observation period, all predictable variables were collected upon study enrollment, because the purpose of the study was to determine the associations of clinical features at enrollment and life prognosis.
The use of non-steroidal anti-inflammatory drugs (NSAIDs) at study enrollment was also collected in the current analysis (no use, current use, or habitual use at enrollment), because ibuprofen can be associated with PD risk reduction [17, 18].
Cross-sectional association of CRP concentrations and PD progression
In addition to survival time analysis, the cross-sectional association between CRP concentrations and PD progression was assessed using a case-control study with data obtained upon study enrollment. After assigning patients who presented with rapid progression as “cases” and others as “controls,” the association was investigated as an odds ratio using a multivariate logistic regression model.
Variables
Age, MMSE, PD duration (years), albumin (mg/dL), and CRP (mg/L) were regarded as scale variables. Due to the non-Gaussian distribution, log2 CRP [2], instead of CRP, was used in multivariable statistic models. Sex (male/female) was regarded as categorical, and mH-Y stage was handled as dichotomous (mH-Y 1–3 vs. 4–5) in the statistical analysis.
Sample size and statistical analysis
Based on a previous study [19], the cumulative survival rate was assumed to be 80% and 60% for the lower two thirds and the top third of CRP concentrations upon study enrollment, respectively. The sample size was estimated at 219 with 90% power and p < 0.05.
Kaplan-Meier curves for the cumulative incidence of death were obtained after dividing patients into two groups according to clinical features. The log-rank test was used to determine the associations between life prognosis and clinical factors. The HRs of baseline CRP concentrations for deaths were estimated using the Cox proportional hazard model after adjusting for age (per 10 years), sex (male or female), PD duration (per 5 years), mH-Y (1–3 or 4–5), serum albumin (per mg/dL), and MMSE (> 24 or ≤ 24). By changing the cut-off value of the baseline CRP concentration, HRs were estimated with 95% confidence intervals. After stratifying patients according to age, sex, PD duration, mH-Y, MMSE, and serum albumin, Kaplan-Meier curves were used to confirm that the association between CRP and life prognosis was independent from these factors. Statistical significance was tested as a pooled p-value of a Log-rank test. P-values < 0.05 were considered statistically significant. All statistical analyses were performed using the statistical software IBM SPSS version 21 and Graph Pad software 5.0.
Results
The CRP concentrations at follow-up remained within the range of ≤ 0.8 mg/L in 77.5%, 75.5%, 75.8%, and 74.1% of patients with CRP ≤ 0.8 mg/L at the first sampling, as well as at the second (21–180 days), third (181–360 days), fourth (361–720 days), and fifth (721–1,080 days) follow-up blood sampling, respectively. Levels remained > 0.8 mg/L in 59.2%, 73.4%, 60.9%, and 56.1% of patients with CRP > 0.8 mg/L at the second, third, fourth, and fifth follow-up blood samplings, respectively. Fig 1 illustrates the chronological changes in baseline CRP concentrations, showing the stability of baseline CRP. Upon study enrollment, the mean CRP concentration was 1.53 mg/L, and after logarithm transformation, the mean log2 CRP was −0.79 ± 1.79 (± SD).
CRP was measured at study enrollment and during the follow-up period (21–180 days, 181–360 days, 361–720 days, and 721–1080 days after study enrollment). Patients were assigned to three groups: those with low-level (< 0.3 mg/L), mid-level (0.3–0.8 mg/L), and high-level (> 0.8 mg/L) CRP at enrollment. Data represent median, and top and bottom error bars represent 75 and 25 percentile values, respectively.
S1 Table summarizes the clinical features of the study patients. Age, mH-Y, and CRP were significantly higher, albumin was lower, and PD duration was longer in those who died compared with those who were still alive at study completion. S1 Fig is a histogram of log2 CRP measured upon study enrollment. The proportion of deceased participants to survivals increased with increasing log2 CRP concentrations.
The main causes of deaths were pneumonia (n = 23, 40.4%), sudden death (n = 11, 19.3%), cancer (n = 6, 10.5%), and suffocation (n = 3, 5.3%). Pneumonia is a well-known cause of death in PD and is associated with swallowing disturbances. In addition to pneumonia, suffocation is thought to be related to PD. In this study, sudden death was likely due to PD-related autonomic failure. One patient committed suicide due to PD-related depression. Another patient drowned in a river while in a state of confusion due to PD psychosis. Refractory hypoglycemia was identified in two patients with no history of insulin use or diabetes mellitus. The patients who died of hypoglycemia were emaciated and no cause of emaciation other than PD was identified. Therefore, death from refractory hypoglycemia was regarded as PD-related death. In contrast, cancer deaths, vascular deaths (stroke and pulmonary thrombosis), heart failure, and hepatic failure were regarded as PD-unrelated deaths.
S2 Fig provides scattered plots of the scale predictable variables. The plots showed no multicollinearity among age, PD duration, MMSE, serum albumin, and log2 CRP, although there was a slight but significant difference in CRP between mH-Y stages (S3 Fig). We generated Kaplan-Meier survival curves to compare life prognosis among the clinical factors. As shown in Fig 2, the cumulative survival rate was better in patients with CRP ≤ 0.8 mg/L than in those with CRP > 0.8 mg/L. Furthermore, the survival rate was significant higher in patients with PD duration ≤ 8 years compared with > 8 years, as well as in those with mH-Y 1–3 compared with mH-Y 4–5, and in those with albumin > 4.0mg/dL compared with ≤ 4.0mg/dL. In contrast, the survival rate was not significantly influenced by MMSE, age, or sex. There was no difference in survival time between NSAID non-users and current or habitual users [2,978 (2,836–3,119) days in non-users and 3,019 (2,698–3,341) days in current or habitual users].
Cumulative survival rates according to CRP levels (A), PD disease duration (B), modified H-Y stages (C), MMSE score (D), age (E), sex (F), NSAID use (G), and serum albumin (H). There was a statistically significant difference in the survival rate between patients with CRP > 0.8 mg/L and ≤ 0.8 mg/L (p = 0.00004) (A), patients with PD disease duration ≤ 8 years and > 8 years (p = 0.00004) (B), those with modified H-Y 1–3 and modified H-Y 4–5 (p = 0.00001) (C), and those with albumin > 4.0mg/dL and albumin ≤ 4.0mg/dL (p = 0.004). However, the survival rate was not influenced by MMSE score, (p = 0.499) (D), age (p = 0.117) (E), sex (p = 0.186) (F), or NSAID use (p = 0.846). Statistical significance was tested using the Log-rank test.
We also calculated the HRs for all deaths according to CRP level using the Cox proportional hazard model. The assumption for proportionality of hazards was met in log minus log plots (data not shown). Unadjusted HRs for all deaths increased with log CRP concentrations. The HRs adjusted for age, sex, PD duration, mH-Y, MMSE, and albumin for all deaths increased with increases in CRP concentration, and the associations between CRP and HRs were almost linear. After regarding PD-unrelated deaths (cancer deaths, vascular deaths, heart failure, and hepatic failure), as well as deaths from unknown causes, as censored due to “alternative outcomes,” the associations were very similar; the unadjusted and adjusted HR increased with log CRP concentrations. The HR values of deaths from pneumonia, sudden deaths, and cancer deaths were also almost linearly associated with CRP (Fig 3). The HR values of log2 CRP for all deaths and PD-related deaths are shown in Table 1.
Relative risk of death was estimated as hazard ratios (HRs) using Cox proportional hazard models. Unadjusted HRs of all deaths correlated with log CRP concentrations (A). To exclude possible confounders, HRs of all deaths were adjusted for age, sex, PD disease duration, MMSE (≤24 vs. >24), mH-Y (1–3 vs. 4–5), and serum albumin levels, and the analysis showed a log-linear association with CRP concentrations (B). Similarly unadjusted and adjusted HRs of death, excluding PD-unrelated deaths, correlated with CRP concentrations (C, D). Unadjusted and age- and sex-adjusted HRs of deaths from pneumonia correlated with CRP concentrations (E, F). Unadjusted and age- and sex-adjusted HRs of sudden deaths correlated with CRP concentrations, although there was no significance (G, H). Unadjusted and age- and sex-adjusted HRs of deaths from cancer correlated with CRP concentration, although there was no significance (I, J).
After stratifying patients by age (≤ 70 vs. > 70 years), sex (males vs. females), PD duration (≤ 8 vs. > 8 years), mH-Y stage (1–3 vs. 4–5), MMSE (> 24 vs. ≤ 24), and albumin (> 4.0 mg/dL vs. ≤ 4.0mg/dL), the Kaplan-Meier survival curves for CRP concentration (≤ 0.8 vs. > 0.8 mg/L) showed significantly higher survival rates for patients with CRP ≤ 0.8 mg/L compared with those with CRP > 0.8 mg/L (p < 0.05, Fig 4).
The cumulative survival rate was compared in patients with CRP ≤ 0.8 mg/L and those with CRP > 0.8 mg/L and was stratified by age (A), sex (B), PD duration (C), modified H-Y stage (D), MMSE (E), and serum albumin (F). The cumulative survival rate was significantly higher in patients with CRP ≤ 0.8 mg/L than those with CRP > 0.8 mg/L. Statistical significance was calculated in pooled models.
Finally, to determine the association of CRP concentration with PD disease progression, we compared CRP concentrations between cases with rapid progression and those with non-rapid progression using data obtained upon study enrollment. Rapid progression was defined as patients with mH-Y of 4–5 and PD duration ≤ 8 years. Multivariate logistic regression analysis that incorporated log2 CRP, age, sex, and MMSE (≤ 24 vs. 24) showed that log2 CRP and age were significantly associated with rapid progression (odds ratio; 1.34 (95% CI 1.12–1.60) per two-fold CRP concentration and 2.11 (95% CI 1.36–3.29) per 10 years of age). Another analysis in which rapid progression was defined as patients with mH-Y 4–5 and PD duration ≤ 5 years, identified log2 CRP, but not age, as a significant determinant of rapid progression (odds ratio; 1.36 (95% CI 1.05–1.75)).
Discussion
The median CRP level in the present study was 0.50 mg/L, with a mean log2 CRP of -0.79, i.e., 0.58 mg/L (2−0.79 = 0.58 mg/L). These values are identical to those reported in our previous cross-sectional study of PD patients [20] and to those reported in a large population-based Japanese study [21], although they are lower than those reported in previous studies from other countries [2]. In contrast to the mean and median values, the standard deviation of log2 CRP (1.79) was similar to that reported in a previous meta-analysis study [2], suggesting a left-shift in the distribution in Japanese compared to other races. The reason for the low CRP in Japanese is unknown at present, but it might be due to genetic background rather than environmental factors, because CRP concentrations in Japanese Americans are also lower than other populations [22].
Our results showed poor life prognosis for PD patients with high baseline CRP values, long disease duration, high mH-Y stages, and lower serum albumin. While no multicollinearity was noted in predictable variables, we could not exclude possible confounders, because most patients with long disease duration also have advanced mH-Y stage and their CRP may be elevated. To exclude possible confounders, we confirmed the associations using Cox proportional hazard models; the obtained adjusted HR was linearly associated with CRP levels, as shown in Fig 3B.
Focusing on deaths from PD-related events, Fig 3B and 3D show the log-linear association of CRP levels with mortality after adjustment for age, sex, disease duration, mH-Y, MMSE, and serum albumin, indicating that the association is independent of age, disease duration, PD severity, cognitive function, and nutritional conditions.
Multivariate analysis by Cox proportional hazard model identified CRP levels, as well as age, mH-Y, and MMSE, as significant determinants of risk of all deaths and deaths from PD-related events (Table 1). These results are consistent with previous studies, which reported the association of PD mortality with severe H-Y stages [15] and poor cognitive function [23, 24].
In the general population, high CRP levels correlate with risk of vascular death [2] and cancer death [3], as well as the development of diabetes mellitus [22]. In this context, CRP was associated with mortality independent of PD. To resolve this issue, we computed the HR values of deaths from PD-related events after censoring patients who died of cancer, vascular events, and other PD-unrelated events. The analysis showed a significant correlation between CRP levels and PD-related deaths, and this correlation is independent of age, sex, disease duration, PD severity, and cognitive function. As shown in Fig 3I and 3J, the HRs of cancer death increased with increasing CRP. However, the HRs are not significant in this cohort, possibly due to the small number of cancer deaths. Two patients died of refractory hypoglycemia and were free of diabetes with no history of insulin use, but they were emaciated. While refractory hypoglycemia can be induced by certain drugs, no such drugs were prescribed. A similar case was previously reported and we assume that emaciation due to PD was associated with hypoglycemia in that case [25]. Although sudden death was categorized as PD-related death, we could not determine whether it was due to coronary insufficiency or PD-related autonomic failure. Although this is one limitation of our study, the high CRP concentrations were associated with mortality from other PD-related events, such as pneumonia. In addition, high CRP levels were also associated with rapid progression of PD. Taken together, these results support the conclusion that CRP levels correlate with prognosis of PD.
As shown in Fig 2, the use of NSAIDs upon study enrollment was not associated with life prognosis in PD, although medications can change during the long-term clinical course and may influence CRP levels or inflammation.
Previous studies have demonstrated neuroinflammation, such as microglia activation [10] and infiltration of CD4 lymphocyte [26], in the brains of PD patients. In PD, dying neurons are thought to release alpha-synuclein, and this in return is taken up by surrounding neurons where it accumulates in Lewy bodies of degenerating neurons [27, 28]. Alpha-synuclein is oxidized and nitrated, thereby inducing neuroinflammation [29, 30], and this latter process is probably mediated by Fcγ receptors [31], which bind to CRP [32]. CRP also increases permeability of the blood brain barrier through Fcγ receptors, further modulating neuroinflammation [33]. It is known that free radicals, including OH and NO, are generated as a byproduct of mitochondrial oxidative phosphorylation during ATP synthesis. In the presence of iron (Fe), these free radicals participate in the Fenton reaction to synthesize peroxynitrite ions involved in α-synuclein nitration and neuroinflammation [34–38]. Hence in addition to CRP, the ratio of nitrated α-synuclein/native α-synuclein may be used as an early and sensitive biomarker of neuro-inflammation in PD [38] (S4 Fig).
High CRP levels increase permeability of the blood brain barrier by binding to the Fcγ receptor [33, 39] and elicit microglia activation in the brain [32]. Animal studies have shown that systemic inflammation causes neurodegeneration through microglia activation [40, 41]. CRP is synthesized in the liver and is then transported to accumulate, or is synthesized in neurons, in neurodegenerative lesions in the brains of Alzheimer’s disease patients [8]. Recent evidence suggests that systemic inflammation contributes to the exacerbation of acute symptoms of chronic neurodegenerative disease [20, 42, 43], suggesting that systemic inflammation seems to promote neurodegeneration. Neuroinflammation can also elicit activation of inflammasome and caspase-1 in the brains of patients with Alzheimer disease [44]. Similar to Alzheimer disease, it is thought that aggregation of alpha-synuclein released from neurons can elicit neuroinflammation by microglia activation and inflammasomes in PD [45]. The present results add support to this hypothesis.
Supporting Information
S1 Fig. Histograms of log2 CRP.
Distribution of log2 CRP was bell-shaped, and the proportion of patients who died during the follow-up (green) to those who are still alive (blue) increased with log2 CRP.
https://doi.org/10.1371/journal.pone.0134118.s001
(TIF)
S2 Fig. Scatter plots of scale predictable variables (age, PD disease duration, MMSE, and log2 CRP).
There was no multicollinearity between these parameters.
https://doi.org/10.1371/journal.pone.0134118.s002
(TIF)
S3 Fig. Relationship between plasma CRP levels and PD severity (mH-Y).
Plasma CRP levels were expressed by box-lots according to mH-Y stages. There was a statistically significant difference in CRP between mH-Y stages (one-way ANOVA, p < 0.0001).
https://doi.org/10.1371/journal.pone.0134118.s003
(TIF)
S4 Fig. Possible mechanism of subclinical systemic inflammation and neurodegeneration in PD.
Subclinical systemic inflammation leads to an increase proinflammatory state, which elicits neuroinflammation and nitration of a-synuclein. This further causes microglia activation, inflammasome formation, and mitochondrial damage.
https://doi.org/10.1371/journal.pone.0134118.s004
(TIF)
S1 Table. Baseline clinical features by life outcomes in Parkinson disease.
https://doi.org/10.1371/journal.pone.0134118.s005
(PDF)
Author Contributions
Conceived and designed the experiments: H. Sawada TO. Performed the experiments: H. Sawada TO AU ST MK KY KP H. Sugiyama. Analyzed the data: H. Sawada TO. Wrote the paper: H. Sawada TO.
References
- 1. Danesh J, Wheeler JG, Hirschfield GM, Eda S, Eiriksdottir G, Rumley A, et al. C-reactive protein and other circulating markers of inflammation in the prediction of coronary heart disease. N Engl J Med. 2004;350(14):1387–97. Epub 2004/04/09. pmid:15070788.
- 2. Emerging Risk Factors C, Kaptoge S, Di Angelantonio E, Lowe G, Pepys MB, Thompson SG, et al. C-reactive protein concentration and risk of coronary heart disease, stroke, and mortality: an individual participant meta-analysis. Lancet. 2010;375(9709):132–40. Epub 2009/12/25. pmid:20031199; PubMed Central PMCID: PMC3162187.
- 3. Heikkila K, Ebrahim S, Lawlor DA. A systematic review of the association between circulating concentrations of C reactive protein and cancer. J Epidemiol Community Health. 2007;61(9):824–33. Epub 2007/08/19. pmid:17699539; PubMed Central PMCID: PMC2703800.
- 4. Emberson JR, Whincup PH, Morris RW, Walker M, Lowe GD, Rumley A. Extent of regression dilution for established and novel coronary risk factors: results from the British Regional Heart Study. Eur J Cardiovasc Prev Rehabil. 2004;11(2):125–34. Epub 2004/06/10. pmid:15187816.
- 5. Perry VH. The influence of systemic inflammation on inflammation in the brain: implications for chronic neurodegenerative disease. Brain Behav Immun. 2004;18(5):407–13. Epub 2004/07/22. pmid:15265532.
- 6. Holmes C, Cunningham C, Zotova E, Woolford J, Dean C, Kerr S, et al. Systemic inflammation and disease progression in Alzheimer disease. Neurology. 2009;73(10):768–74. Epub 2009/09/10. pmid:19738171; PubMed Central PMCID: PMC2848584.
- 7. Holmes C, Cunningham C, Zotova E, Culliford D, Perry VH. Proinflammatory cytokines, sickness behavior, and Alzheimer disease. Neurology. 2011;77(3):212–8. Epub 2011/07/15. pmid:21753171; PubMed Central PMCID: PMC3136056.
- 8. Yasojima K, Schwab C, McGeer EG, McGeer PL. Human neurons generate C-reactive protein and amyloid P: upregulation in Alzheimer's disease. Brain Res. 2000;887(1):80–9. Epub 2001/01/03. pmid:11134592.
- 9. Hirsch EC, Hunot S. Neuroinflammation in Parkinson's disease: a target for neuroprotection? Lancet Neurol. 2009;8(4):382–97. Epub 2009/03/20. pmid:19296921.
- 10. McGeer PL, Itagaki S, Boyes BE, McGeer EG. Reactive microglia are positive for HLA-DR in the substantia nigra of Parkinson's and Alzheimer's disease brains. Neurology. 1988;38(8):1285–91. Epub 1988/08/01. pmid:3399080.
- 11. Imamura K, Hishikawa N, Sawada M, Nagatsu T, Yoshida M, Hashizume Y. Distribution of major histocompatibility complex class II-positive microglia and cytokine profile of Parkinson's disease brains. Acta Neuropathol. 2003;106(6):518–26. Epub 2003/09/27. pmid:14513261.
- 12. Gerhard A, Pavese N, Hotton G, Turkheimer F, Es M, Hammers A, et al. In vivo imaging of microglial activation with [11C](R)-PK11195 PET in idiopathic Parkinson's disease. Neurobiol Dis. 2006;21(2):404–12. Epub 2005/09/27. pmid:16182554.
- 13. Mortensen RF, Zhong W. Regulation of phagocytic leukocyte activities by C-reactive protein. J Leukoc Biol. 2000;67(4):495–500. Epub 2000/04/19. pmid:10770281.
- 14. Perry VH, Cunningham C, Holmes C. Systemic infections and inflammation affect chronic neurodegeneration. Nat Rev Immunol. 2007;7(2):161–7. Epub 2007/01/16. pmid:17220915.
- 15. Morgante L, Salemi G, Meneghini F, Di Rosa AE, Epifanio A, Grigoletto F, et al. Parkinson disease survival: a population-based study. Arch Neurol. 2000;57(4):507–12. Epub 2000/04/18. pmid:10768625.
- 16. Matsumoto H, Sengoku R, Saito Y, Kakuta Y, Murayama S, Imafuku I. Sudden death in Parkinson's disease: a retrospective autopsy study. J Neurol Sci. 2014;343(1–2):149–52. Epub 2014/06/15. pmid:24928079.
- 17. Driver JA, Logroscino G, Lu L, Gaziano JM, Kurth T. Use of non-steroidal anti-inflammatory drugs and risk of Parkinson's disease: nested case-control study. BMJ. 2011;342:d198. Epub 2011/01/22. pmid:21252104; PubMed Central PMCID: PMC3023971.
- 18. Gao X, Chen H, Schwarzschild MA, Ascherio A. Use of ibuprofen and risk of Parkinson disease. Neurology. 2011;76(10):863–9. Epub 2011/03/04. pmid:21368281; PubMed Central PMCID: PMC3059148.
- 19. Ben-Shlomo Y, Marmot MG. Survival and cause of death in a cohort of patients with parkinsonism: possible clues to aetiology? J Neurol Neurosurg Psychiatry. 1995;58(3):293–9. Epub 1995/03/01. pmid:7897409; PubMed Central PMCID: PMC1073364.
- 20. Sawada H, Oeda T, Umemura A, Tomita S, Hayashi R, Kohsaka M, et al. Subclinical elevation of plasma C-reactive protein and illusions/hallucinations in subjects with Parkinson's disease: case-control study. PLoS One. 2014;9(1):e85886. Epub 2014/02/06. pmid:24497930; PubMed Central PMCID: PMC3908859.
- 21. Arima H, Kubo M, Yonemoto K, Doi Y, Ninomiya T, Tanizaki Y, et al. High-sensitivity C-reactive protein and coronary heart disease in a general population of Japanese: the Hisayama study. Arteriosclerosis, thrombosis, and vascular biology. 2008;28(7):1385–91. Epub 2008/04/12. pmid:18403728.
- 22. Nakanishi S, Yamane K, Kamei N, Okubo M, Kohno N. Elevated C-reactive protein is a risk factor for the development of type 2 diabetes in Japanese Americans. Diabetes care. 2003;26(10):2754–7. Epub 2003/09/30. pmid:14514575.
- 23. Levy G, Tang MX, Louis ED, Cote LJ, Alfaro B, Mejia H, et al. The association of incident dementia with mortality in PD. Neurology. 2002;59(11):1708–13. Epub 2002/12/11. pmid:12473757.
- 24. de Lau LM, Breteler MM. Epidemiology of Parkinson's disease. Lancet Neurol. 2006;5(6):525–35. Epub 2006/05/23. pmid:16713924.
- 25. Shimizu K, Ogura H, Wasa M, Hirose T, Shimazu T, Nagasaka H, et al. Refractory hypoglycemia and subsequent cardiogenic shock in starvation and refeeding: Report of three cases. Nutrition. 2014;30(9):1090–2. Epub 2014/06/15. pmid:24927630.
- 26. Brochard V, Combadiere B, Prigent A, Laouar Y, Perrin A, Beray-Berthat V, et al. Infiltration of CD4+ lymphocytes into the brain contributes to neurodegeneration in a mouse model of Parkinson disease. J Clin Invest. 2009;119(1):182–92. Epub 2008/12/24. pmid:19104149; PubMed Central PMCID: PMC2613467.
- 27. Desplats P, Lee HJ, Bae EJ, Patrick C, Rockenstein E, Crews L, et al. Inclusion formation and neuronal cell death through neuron-to-neuron transmission of alpha-synuclein. Proc Natl Acad Sci U S A. 2009;106(31):13010–5. Epub 2009/08/05. pmid:19651612; PubMed Central PMCID: PMC2722313.
- 28. Emmanouilidou E, Melachroinou K, Roumeliotis T, Garbis SD, Ntzouni M, Margaritis LH, et al. Cell-produced alpha-synuclein is secreted in a calcium-dependent manner by exosomes and impacts neuronal survival. J Neurosci. 2010;30(20):6838–51. Epub 2010/05/21. pmid:20484626; PubMed Central PMCID: PMC3842464.
- 29. Gao HM, Kotzbauer PT, Uryu K, Leight S, Trojanowski JQ, Lee VM. Neuroinflammation and oxidation/nitration of alpha-synuclein linked to dopaminergic neurodegeneration. J Neurosci. 2008;28(30):7687–98. Epub 2008/07/25. pmid:18650345; PubMed Central PMCID: PMC2702093.
- 30. Reynolds AD, Glanzer JG, Kadiu I, Ricardo-Dukelow M, Chaudhuri A, Ciborowski P, et al. Nitrated alpha-synuclein-activated microglial profiling for Parkinson's disease. J Neurochem. 2008;104(6):1504–25. Epub 2007/11/27. pmid:18036154.
- 31. Cao S, Standaert DG, Harms AS. The gamma chain subunit of Fc receptors is required for alpha-synuclein-induced pro-inflammatory signaling in microglia. J Neuroinflammation. 2012;9:259. Epub 2012/11/29. pmid:23186369; PubMed Central PMCID: PMC3526448.
- 32. Li YN, Qin XJ, Kuang F, Wu R, Duan XL, Ju G, et al. Alterations of Fc gamma receptor I and Toll-like receptor 4 mediate the antiinflammatory actions of microglia and astrocytes after adrenaline-induced blood-brain barrier opening in rats. J Neurosci Res. 2008;86(16):3556–65. Epub 2008/08/30. pmid:18756515.
- 33. Closhen D, Bender B, Luhmann HJ, Kuhlmann CR. CRP-induced levels of oxidative stress are higher in brain than aortic endothelial cells. Cytokine. 2010;50(2):117–20. Epub 2010/03/09. pmid:20207160.
- 34. Ibi M, Sawada H, Kume T, Katsuki H, Kaneko S, Shimohama S, et al. Depletion of intracellular glutathione increases susceptibility to nitric oxide in mesencephalic dopaminergic neurons. J Neuroch. 1999;73(4):1696–703. Epub 1999/09/29. pmid:10501217.
- 35. Sawada H, Kawamura T, Shimohama S, Akaike A, Kimura J. Different mechanisms of glutamate-induced neuronal death between dopaminergic and non-dopaminergic neurons in rat mesencephalic culture. J Neurosci Res. 1996;43(4):503–10. Epub 1996/02/15. pmid:8699537.
- 36. Ebadi M, Sharma SK. Peroxynitrite and mitochondrial dysfunction in the pathogenesis of Parkinson's disease. Antioxidants & redox signaling. 2003;5(3):319–35. Epub 2003/07/26. pmid:12880486.
- 37. Sawada H, Shimohama S, Tamura Y, Kawamura T, Akaike A, Kimura J. Methylphenylpyridium ion (MPP+) enhances glutamate-induced cytotoxicity against dopaminergic neurons in cultured rat mesencephalon. J Neurosci Res. 1996;43(1):55–62. Epub 1996/01/01. pmid:8838574.
- 38. Sharma SK, Ebadi M. Metallothionein attenuates 3-morpholinosydnonimine (SIN-1)-induced oxidative stress in dopaminergic neurons. Antioxidants & redox signaling. 2003;5(3):251–64. Epub 2003/07/26. pmid:12880480.
- 39. Kuhlmann CR, Librizzi L, Closhen D, Pflanzner T, Lessmann V, Pietrzik CU, et al. Mechanisms of C-reactive protein-induced blood-brain barrier disruption. Stroke. 2009;40(4):1458–66. Epub 2009/02/28. pmid:19246692.
- 40. Lunnon K, Teeling JL, Tutt AL, Cragg MS, Glennie MJ, Perry VH. Systemic inflammation modulates Fc receptor expression on microglia during chronic neurodegeneration. J Immunol. 2011;186(12):7215–24. Epub 2011/05/17. pmid:21572034.
- 41. Cerejeira J, Firmino H, Vaz-Serra A, Mukaetova-Ladinska EB. The neuroinflammatory hypothesis of delirium. Acta Neuropathol. 2010;119(6):737–54. Epub 2010/03/24. pmid:20309566.
- 42. Perry VH. Contribution of systemic inflammation to chronic neurodegeneration. Acta Neuropathol. 2010;120(3):277–86. Epub 2010/07/21. pmid:20644946.
- 43. Umemura A, Oeda T, Tomita S, Hayashi R, Kohsaka M, Park K, et al. Delirium and high Fever are associated with subacute motor deterioration in Parkinson disease: a nested case-control study. PLoS One. 2014;9(6):e94944. Epub 2014/06/03. pmid:24887491; PubMed Central PMCID: PMC4041721.
- 44. Heneka MT, Kummer MP, Stutz A, Delekate A, Schwartz S, Vieira-Saecker A, et al. NLRP3 is activated in Alzheimer's disease and contributes to pathology in APP/PS1 mice. Nature. 2013;493(7434):674–8. Epub 2012/12/21. pmid:23254930; PubMed Central PMCID: PMC3812809.
- 45. Codolo G, Plotegher N, Pozzobon T, Brucale M, Tessari I, Bubacco L, et al. Triggering of inflammasome by aggregated alpha-synuclein, an inflammatory response in synucleinopathies. PLoS One. 2013;8(1):e55375. Epub 2013/02/06. pmid:23383169; PubMed Central PMCID: PMC3561263.