Edited by: Alain Kaelin-Lang, Neurocenter of Southern Switzerland, Switzerland
Reviewed by: Chien Tai Hong, Taipei Medical University, Taiwan; Zhong Pei, Sun Yat-sen University, China
This article was submitted to Movement Disorders, a section of the journal Frontiers in Neurology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Parkinson's disease (PD) is a severe neurodegenerative disease, affecting 1% of the population above the age of 60 (
Clinically, disrupted lipid metabolism can lead to abnormal serum lipid levels of TC, LDL-C, HDL-C, and TG. However, the causal relationship between serum lipid levels and the risk of PD has not yet been determined. Since the first attempt in 1994 (
Although a meta-analysis in 2013 observed no association between serum cholesterol and the risk of PD (
Our meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement (see
Studies were included if they fulfilled the following criteria: (
Two authors (ZJ and XRX) independently extracted the following data from included studies: first author and publication year; study location; study design; cohort name; serum lipid parameters; data type; follow-up time; number of participants/person-years, number of cases(overall and for each category); age; outcome definition (relative risk[RR]/odds ratio[OR]/hazard ratio[HR]); covariates adjusted; levels of exposure, along with corresponding risk estimates and 95% CI for each category. Disagreements were settled by mutual discussion. If articles selected in this paper report concentrations of serum lipids by the International System of Units (SI), we converted those to conventional units with a conversion ratio of 38.6 (1 mg/dL = 0.0259 mmol/L) for TC, LDL-C, and HDL-C or 88.5 (1 mg/dL = 0.0113 mmol/L) for TG.
The Newcastle-Ottawa Scale (NOS) was used to evaluate the quality of the included observational studies (
Because PD is a relatively uncommon disease, HR or OR is considered to mathematically approximate RR (
A dose-response meta-analysis was also performed to explore the dose-risk relationship between exposure and risk. The category serum lipid levels were determined as the mean or median value in each category if available for each study. For studies only gave the range in a category, the midpoint value was assigned for closed categories. In the case of open-ended highest or lowest category, the category serum lipid levels were equal to the lower or upper boundary plus or minus 1.5 folds the range of the closest category. When studies set the highest categories as the reference groups, we uniformly converted them to the lowest categories as needed (
All statistical analyses were performed using the software of Stata 12.0 (StataCorp, Texas, USA).
Detailed information on the literature search could be seen from
The flow chart for detailed steps of the literature search.
These 15 studies were published between 1994 and 2019. Three of the studies were performed primarily in Americas (
Quality assessment of each eligible study according to the NOS.
Rozani (2018) | * | * | * | * | ** | * | * | * |
Nam (2018) | * | * | * | * | * | * | * | * |
Huang (2015) | * | * | * | ** | * | * | * | |
Hu (2008) | * | * | * | * | * | * | * | * |
Friedman (2013) | * | * | * | * | ** | * | * | |
Simon (2007) | * | * | * | * | * | * | ||
Saaksjarvi (2015) | * | * | * | * | * | * | * | * |
Huang (2008) | * | * | * | * | * | * | * | |
Grandinetti (1994) | * | * | * | * | * | * | ||
De Lau (2006) | * | * | * | * | * | * | * | * |
Benn (2017) | * | * | * | * | * | * | * | * |
Jeong (2019) | * | * | * | * | ** | * | * | * |
Vikdahl (2015) | * | * | * | * | * | * | * | * |
Miyake (2010) | * | * | * | * | * | |||
Savica (2012) | * | * | * | * | * | * | * |
For the TC group, eight cohort studies (
Forest plot presenting the pooled estimate effects (RR) on the relationship between serum TC levels and risk of PD. RR, relative risk; CI, confidence interval; TC, total cholesterol; PD, Parkinson's disease.
In subgroup analyses by study location, study design, statin use, quality of study, and gender, overall pooled risk estimate remained non-significant, and no evidence of between-study heterogeneity was found (
Subgroup analyses (serum TC levels and PD risk).
Europe | 5 | 0.95(0.66–1.37) | 0.794 | 80.8 | 0.000 | 0.690 |
Americas | 3 | 0.87(0.50–1.52) | 0.631 | 70.3 | 0.034 | |
Asia | 3 | 0.83(0.49–1.41) | 0.492 | 59.1 | 0.087 | |
cohort | 8 | 0.90(0.67–1.21) | 0.485 | 75.7 | 0.000 | 0.973 |
case-control | 3 | 0.91(0.60–1.37) | 0.635 | 57.0 | 0.098 | |
YES | 6 | 1.01(0.65–1.57) | 0.970 | 81.6 | 0.000 | 0.374 |
NO | 5 | 0.87(0.71–1.05) | 0.151 | 33.1 | 0.201 | |
≥7 | 8 | 0.96(0.70–1.32) | 0.810 | 76.1 | 0.000 | 0.467 |
<7 | 3 | 0.81(0.59–1.12) | 0.201 | 49.4 | 0.139 | |
Male | 7 | 0.87(0.63–1.21) | 0.414 | 66.8 | 0.006 | 0.918 |
Female | 6 | 0.90(0.56–1.43) | 0.649 | 71.3 | 0.004 |
The funnel plot was visually asymmetry (
Funnel plot of serum TC levels and risk of PD. RR, relative risk; TC, total cholesterol; PD, Parkinson's disease.
Five cohort studies (
Dose-response relationship between serum TC levels and risk of PD. TC, total cholesterol; PD, Parkinson's disease.
For the LDL-C group, five cohort studies (
Forest plot presenting the pooled estimate effects(RR) on the relationship between serum LDL-C levels and risk of PD. RR, relative risk; CI, confidence interval; LDL-C, low-density lipoprotein cholesterol; PD, Parkinson's disease.
In subgroup analyses according to study location, statin use, and gender, similar significant inverse relationships between serum LDL-C and PD risk were observed for studies confined to a location in Americas (RR 0.43, 95% CI 0.21–0.88), statin adjustment (RR 0.66, 95% CI 0.49–0.90), or male (RR 0.69, 95% CI 0.54–0.88). Besides, the substantial heterogeneity disappeared in the male group. We summarized the detailed results in
Subgroup analyses (serum LDL-C levels and PD risk).
Europe | 3 | 0.82 (0.64–1.04) | 0.099 | 50.3 | 0.134 |
Americas | 1 | 0.43 (0.21–0.88) | 0.021 | — | — |
Asia | 1 | 0.60 (0.40–1.10) | 0.099 | — | — |
YES | 3 | 0.66 (0.49–0.90) | 0.006 | 35.3 | 0.213 |
NO | 2 | 0.81 (0.50–1.31) | 0.085 | 66.9 | 0.082 |
Male | 2 | 0.69 (0.54–0.88) | 0.003 | 0 | 0.537 |
Female | 1 | 0.88 (0.62–1.29) | 0.494 | — | — |
Both Egger's test (
Four cohort studies (
Dose-response relationship between serum LDL-C levels and risk of PD. LDL-C, low-density lipoprotein cholesterol; PD, Parkinson's disease.
For the HDL-C group, we included five cohort studies (
Forest plot presenting the pooled estimate effects(RR) on the relationship between serum HDL-C levels and risk of PD. RR, relative risk; CI, confidence interval; HDL-C, high-density lipoprotein cholesterol; PD, Parkinson's disease.
As all included studies were of high quality, we conducted subgroup analyses following study location, statin use, and gender. The pooled risk estimate remained non-significant, and between-study heterogeneity decreased in either Europe or the male group (
There was no indication of publication bias assessed with Egger's test (
For the TG group, we obtained three cohort studies (
Forest plot presenting the pooled estimate effects(RR) on the relationship between serum TG levels and risk of PD. RR, relative risk; CI, confidence interval; TG, triglycerides; PD, Parkinson's disease.
Sensitivity analyses indicated little change in the pooled RR when any study was removed (data not shown).
In this meta-analysis, we summarized the current evidence on the association between serum lipid parameters and the risk of PD, and thus provides a unique perspective to evaluate the effects of different serum lipid parameters on PD risk. There are some notable strengths deserving mention. Firstly, four serum lipid parameters, including TC, LDL-C, HDL-C, and TG, were included in a quantitative review to investigate their effects on PD risk for the first time. Secondly, the strict inclusion criterion of providing information on serum lipid levels before PD onset avoided the possibility of reverse causality from abnormal lipid profile induced by PD. Thirdly, both the high vs. low analysis and dose-response analysis were conducted, which ensures the reliability of the outcome.
In the high vs. low analysis, we found that the exposure of higher serum LDL-C levels may decrease the future risk of PD, which is consistent with the previous data on the association between serum LDL-C and PD risk (
Multiple lines of biological evidence responsible for the inverse linkage between serum cholesterol and PD risk have been found. Some rodent experiments supported the protective role of cholesterol precursors in the pathogenesis of PD. For example, squalene might save the striatal neurons from toxic effects caused by 6-hydroxydopamine (
In contrast, some hypotheses support a positive relationship between serum cholesterol levels and PD risk. As is generally recognized, abnormal intracellular aggregation of α-synuclein has an etiological role in the occurrence and development of PD (
While exploring serum lipids in relation to the risk of PD, we should draw more attention to the confounding effect of lipid-lowering drugs on this association. Statins are cholesterol-lowering drugs widely prescribed to prevent and treat cardiovascular and cerebrovascular diseases. Some studies indicated that statins might have anti-inflammation activity (
For this paper, we should consider several limitations. Firstly, as mentioned before, altered serum lipid levels might represent a prodromal symptom of PD. Without taking into consideration the duration of the prodromal period, while still challenging to ascertain, we include the studies using data of serum lipid levels measured just before the onset of PD, which might make the results somewhat biased. Secondly, a limited number of the included studies for the relationship between serum TG and PD risk urges us to interpret its negative results with caution. Thirdly, there are some missing data in the dose-response analysis, and the approximate values inferred by existing methods may not be accurate, so the results of the dose-risk relationship might demand more sufficient data.
Our meta-analysis showed that higher serum LDL-C might be a protective factor for PD and somehow reduce the future risk of PD. From an overall perspective, the protective effect of high serum LDL-C on PD risk attends to trifles and neglects essentials. After all, the increasing threat of cardiovascular and cerebrovascular events induced by hypercholesterolemia is a major public health concern. However, this study still has some implications for exploring the role of cholesterol in the pathogenesis of PD.
All datasets generated for this study are included in the article/
ZJ, XX, and RO designed the study and did the literature review. ZJ did the statistical analysis, edited tables and pictures, and wrote the primary manuscript. XG contributed to revising the grammar. XL helped to promote the methodology. HS and WS reviewed the final manuscript. All authors contributed to the article and approved the submitted version.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The Supplementary Material for this article can be found online at: