The role of serum lipids in Parkinson’s disease (PD) remains controversial. We aimed to evaluate the association between time-varying serum lipid levels and the risk of PD. This study included an assessment of the complete lipid profiles of 200,454 individuals from the 2002–2019 Korean National Health Insurance Health Screening Cohort. Time-dependent Cox proportional hazard regression models were used to evaluate the association between serum lipid levels over time and the risk of PD. Individuals in the lowest tertile of total cholesterol and low-density lipoprotein cholesterol had a 1.17 times [hazard ratio (HR) 1.17; 95% confidence interval (CI) 1.04–1.31] and 1.19 times (HR 1.19; 95% CI 1.06–1.34) higher risk of PD than those in middle tertile, respectively. Individuals in the highest high-density lipoprotein cholesterol tertile had a 0.89 times (HR 0.89; 95% CI 0.79–1.00) lower risk of PD than those in middle tertile, but the association was less robust in sensitivity analyses. Serum triglyceride levels were not related to the risk of PD. Our results suggest that the serum total and low-density lipoprotein cholesterol levels over time are inversely associated with the risk of PD. Further research is warranted to confirm these findings and reveal the underlying mechanisms.
Parkinson’s disease (PD) is the second most common neurodegenerative disease, with a lifetime risk of 2% among men and 1.3% among women1.
The pathological features of PD include the degeneration of dopaminergic neurons in the substantia nigra and widespread accumulation of α-synuclein2,3. Although the etiology of PD is not fully understood, epidemiologic studies have identified potential risk factors and protective factors for PD, including serum lipid levels1. However, studies investigating the association between serum lipid levels and PD risk have yielded inconsistent findings. A Finnish cohort study reported that individuals with higher serum total cholesterol (TC) levels had an increased risk of PD than those with lower TC level4. In a Korean cohort study, higher triglyceride (TG) and high-density lipoprotein cholesterol (HDL-C) levels were associated with an increased risk of PD5.
The risk of PD was not associated with history of diagnosed hypercholesterolemia in the Nurses’ Health Study and Health Professionals Follow-Up Study 6. In a Danish study using the Mendelian randomization method, no association was found between low-density lipoprotein cholesterol (LDL-C) and the risk of PD, whereas a Finnish study using a similar method reported that high levels of TC, TG, and LDL-C were related to a lower risk of PD7,8.
Other recent cohort and case–control studies reported that TC, LDL-C, HDL-C, and TG levels were inversely associated with the risk of PD9,10,11,12,13,14.
Heterogeneity among studies may be due to differences in methodology, including study design, sample size, and control for possible confounders. Although many studies had a prospective design, most did not reflect the time-varying nature of serum lipid levels over time. In addition, studies often overlooked the effect of potential confounders, such as statin use, lifestyle, and medical history, which were also thought to be related to the risk of PD. Finally, there is limited longitudinal evidence on the relationship between blood lipid levels and the risk of PD in Asian populations.
Therefore, this study aimed to investigate the association between time-varying serum lipid levels (TC, LDL-C, HDL-C, and TG) and the incidence of PD using Korean population-based data.
Between January 1, 2002, and December 31, 2019, 257,943 individuals met the inclusion criteria for our study. Individuals with a medical record of PD or secondary Parkinsonism before or during the first year of follow-up (n = 1,349), those who died during the first year of follow-up (n = 1698), and those with statin prescriptions (n = 54,442) prior to the first health screening were excluded. The final sample included 200,454 individuals (Fig. 1).
The mean and cumulative follow-up time were 8.5 years and 1,713,530 person-years, respectively, and 1712 PD cases were identified during the study period. The mean age of the participants at baseline was 57.4 (± 8.6) years and 113,577 (56.7%) of the participants were male. The mean serum lipid levels at baseline were 198.7 (± 34.4) mg/dL for TC, 118.6 (± 35.6) mg/dL for LDL-C, 54.8 (± 23.9) mg/dL for HDL-C, and 131.4 (± 81.6) mg/dL for TG. A total of 41,851 (20.9%) individuals started statins after enrollment in the study (Table 1).
The crude incidence rates of PD (per 100,000 person-years) among individuals in the lowest, middle, and highest tertiles of each serum lipid fraction were 128.9, 88.3, and 80.4 for TC, 125.3. 89.7, and 83.9 for LDL-C, 115.5, 103.8, and 83.9 for HDL-C, 100.7, 105.6, and 93.1 for TG, respectively. After adjusting for all covariates, individuals in the lowest TC tertile had a 1.17 times [hazard ration (HR) 1.17; 95% confidence interval (CI) 1.04–1.31] higher risk of PD than those in the middle TC tertile. Individuals in the highest TC tertile showed no statistically significant difference (HR 1.03; 95% CI 0.91–1.17) in the risk of PD, compared with those in the middle TC tertiles. Compared to individuals with middle LDL-C tertile, those with lowest LDL-C tertile had 1.19 times (HR 1.19; 95% CI 1.06–1.34) higher risk of PD, after adjusting all covariates. Individuals in the highest and middle LDL-C tertiles did not show a difference in the risk of PD (HR 1.06; 95% CI 0.94–1.20). Compared to individuals in the middle HDL-C tertile, those in the lowest HDL-C tertile did not show a difference in the risk of PD (HR 1.01; 95% CI 0.90–1.14), while those in the highest HDL-C tertile had a 0.89 times (HR 0.89; 95% CI 0.79–1.00) lower risk of PD. Serum TG level was not associated with the risk of developing PD (Table 2).
Sensitivity analyses showed results similar to those of the main analysis of serum TC and LDL-C levels. However, only two of the four sensitivity analyses showed statistical significance for serum HDL-C levels, although the directions of association were the same as those in the main analysis (Tables 3 and 4). Among 1712 patients with PD, the risk of dementia in PD was related to serum TC, HDL-C, and TG levels (although the directions of the association were not uniform) whereas PD mortality was not associated with any serum lipid level (Supplementary Table S1 and S2).
IN this population-based cohort study, we found that lower serum TC and LDL-C levels were associated with an increased risk of PD, whereas higher serum HDL-C levels were associated with a decreased risk of PD, even after adjusting for potential confounders, including statin use. However, serum TG levels were not related to PD risk in our study. In the sensitivity analyses, the association between serum HDL-C level and PD was less robust, while other findings were similar to those of the main analysis.
Our results are consistent with findings from previous studies—serum TC, LDL-C, and HDL-C were inversely associated with the risk of PD8,9,13; individuals with low cholesterol intake had an increased risk of PD15,16, and high TC levels were related to the slow progression of PD symptoms17. Our findings provide longitudinal evidence for the association between serum lipid levels and PD risk in the Asian population. Moreover, we assessed repeated measurements of serum lipid levels over time and performed sensitivity analyses to assess the robustness of the relationship between serum lipid levels and PD risk.
In the present study, the impact of serum cholesterol levels on the risk of PD was small or modest compared to previous studies that reported that the risk of PD differed by up to two-fold according to the serum cholesterol level11,13. Potential confounders, such as lifestyle factors, comorbidities, and statin use, which were often neglected in other studies, might have contributed to this gap. Furthermore, we found that an altered serum lipid profile may be related to the development of dementia after PD diagnosis, although the findings were preliminary. The lack of association between PD-mortality and serum lipid levels might imply that an altered lipid profile was possibly related to a cognitive function independently of other conditions, such as nutritional status.
The underlying mechanisms of the findings of the present study are unclear and require further research. Elevated cholesterol levels might be related to protective factors of PD (e.g., increased serum cholesterol due to smoking and physical activity after dyslipidemia diagnosis)1. In addition, neurodegenerative changes prior to the diagnosis of PD may have led to lifestyle changes that are linked to cholesterol changes.
As suggested in literature, dysregulation of cholesterol metabolism may be related to neurodegenerative processes in the brain8,9. Cholesterol is an essential component of cell membranes and myelin sheaths, and maintaining balanced cholesterol regulation is important for neuronal signaling and synaptic function18,19,20. The brain contains the highest levels of cholesterol in the body, and alterations in cholesterol homeostasis and biosynthesis can cause neurodegeneration in the central nervous system (CNS)21. In an animal model, mice with cholesterol synthesis deficiency developed motor symptoms, such as tremor and ataxia22. Epidemiological studies have found that abnormal serum cholesterol levels are also associated with other neurodegenerative disorders such as Alzheimer’s disease (AD), Huntington’s disease (HD), and amyotrophic lateral sclerosis23,24,25,26.
However, since it is difficult for lipoprotein-bound cholesterol to cross the blood–brain barrier (BBB), it is difficult to say that alterations in serum cholesterol levels directly affect cholesterol levels in CNS27,28.
Recent studies have found a shared genetic etiology between serum cholesterol levels and PD, as reported in other neurodegenerative disease such as AD and HD25,26,29. Analyses of genome-wide association studies data revealed that dysregulation in lipid metabolism is associated with molecular processes that are considered important in PD development, such as oxidative stress, endoplasmic reticulum stress response, and immune response29,30,31. Combined with the difficulty of lipoprotein crossing the BBB, these findings suggest that serum cholesterol levels and incidence of PD may be indirectly related through genetic pathways.
Our study had several limitations. First, PD diagnosis could be imprecise because we identified PD cases using the ICD-10 codes. Therefore, we tried to improve diagnostic accuracy by defining PD cases as those with at least one admission or two outpatient visits for PD, and performed sensitivity analysis with alternative PD definitions. However, there was still a possibility of missing or misdiagnosed cases due to lack of clinical case certification and individual chart review. The number of patients with PD might have been underestimated, because our PD definition (including sensitivity analyses) did not include the patients with dystonia or other movement disorders (ICD-10 codes G24 and G25), who might have actually had idiopathic PD. Second, we were unable to classify patients with PD genetically (mutations in LRRK2, GBA, or Parkin) due to lack of relevant data. Third, although we adjusted for several potential confounders, residual confounding could be present because information on environmental exposure, genetic factors, dietary factors, and caffeine intake were not available in our data. Fourth, because PD has a long clinical course, and the onset of PD symptoms may precede the diagnosis of PD, the possibility of reverse causation (e.g., low cholesterol levels as a result of PD progression) cannot be excluded. However, our study had a relatively long washout period (minimum 7 years), and our setting with 1-year and 3-year time lags could reduce the possibility of reverse causation. Therefore, future studies including detailed clinical and genetic information on PD are required to verify the findings of the present study.
In conclusion, based on time-varying serum lipids levels, our study suggests that serum TC, LDL-C levels may be inversely associated with the risk of PD after controlling potential confounders. However, further studies are warranted to elucidate underlying mechanisms and specify the optimal range of serum cholesterol levels to minimize the risk of PD.
Study subjects and data sources
Data for this study were obtained from the 2002–2019 Korea National Health Insurance Service Health Screening Cohort (NHIS-HEALS). The National Health Insurance Service (NHIS), a single-payer universal health insurance program, provides a general health-screening program for all Korean citizens aged ≥ 40 years every 2 years32. In 2009, the Korean general health screening program started measuring the full lipid profiles (TC, LDL-C, HDL-C, and TG) of participants. The NHIS-HEALS comprised of 514,866 participants in a general health-screening program, aged between 40 and 79 years in 2002, which is a 10% simple random sample of the target population32. The NHIS-HEALS includes anonymized participant information (demographics, medical records, and health screening database), and the participants were followed up until their loss of eligibility due to death or emigration33.
To fully address time-varying serum lipid levels, only individuals with ≤ 1 missing schedule of biennial health screening were included in the study. Individuals who used statins before study enrollment were excluded to eliminate the potential confounding effects of statins.
This study was approved by the Institutional Review Board of Yonsei University Health System (IRB No: 4-2022-0863) and adhered to the tenets of the Declaration of Helsinki. Written informed consent was waived by Institutional Review Board of Yonsei University Health System (IRB No: 4-2022-0863) because of the retrospective nature of the study and NHIS-HEALS contains anonymized data.
Incident PD cases, the outcome of study, were identified based on the International Statistical Classification of Diseases (ICD-10) code of PD (G20). To reduce false-positive cases and ensure diagnostic validity, individuals with at least two outpatient visits or one admission for PD were classified as patients with PD. Individuals with a medical record of PD or secondary Parkinsonism before the index date were excluded from the study, allowing a minimum washout period of 7 years (2002–2008).
Measurement of lipid profiles and covariates
The Korean general health-screening program has been measuring the serum lipid levels (TC, LDL-C, HDL-C, and TG) of participants every two years since 2009. Serum lipid levels were categorized into tertiles and were included as time-dependent variables. The cutoff values for categorizing each lipid profile were based on values of the first measurements (mg/dL, TC, lowest ≤ 182, middle 183–210, highest ≥ 211; LDL-C, lowest ≤ 103, middle 104–129, highest ≥ 130; HDL-C lowest ≤ 45, middle 46–57, highest ≥ 58; TG low ≤ 88, middle 88–138, highest ≥ 139).
Although we excluded individuals who had used statins before the study period, some participants initiated statin therapy during the study period. Thus, the cumulative duration of statin use (none, < 1 year, 1–2 years, > 2 years) corresponding to the lipid-level-based time segment was included as a covariate. Other potential confounders were identified during study enrollment. Age (continuous variable),sex, area of residence (urban or rural), smoking status (never smoker, ex-smoker, current smoker, or unspecified), alcohol consumption (none, ≤ 7, 7–14, or > 14 units per week), levels of physical activity (0–499, 500–999, or > 1000 MET∙ min per week), body mass index (< 25, ≥ 25 kg/m2, or unspecified), and past medical history (stroke, diabetes mellitus, and hypertension, individuals with more than two outpatient visits or one admission over three years before the index date, based on ICD-10 codes) were included as covariates.
Four additional sensitivity analyses were performed. First, the definition of PD was changed to include participants who used antiparkinsonian drugs (levodopa, levodopa combined with decarboxylase inhibitor or catechol-O-methyl transferase inhibitor, dopamine agonists, monoamine oxidase B inhibitors, amantadine, and anticholinergics) with one or more ICD-10 codes. Second, we changed the definition of PD to include individuals with ≥ 1 hospital admission or ≥ 2 ambulatory visits for PD, and ≥ 1 PD diagnosis made by a neurologist. Third, to assess potential underestimation of PD cases, we used extended definition of PD to include individuals with other forms of secondary parkinsonism (ICD-10 codes G21 and G22), other degenerative diseases in basal ganglia (ICD-10 code G23), and movement disorders in diseases classified elsewhere (ICD-10 code G26). Fourth, the 1-year time lag was replaced by a 3-year time lag to reflect the PD diagnosis.
The baseline characteristics of participants were described as mean and standard deviation for continuous variables, and the number and percentage of participants for categorical variables. To reflect the time from PD onset to diagnosis and reduce the possibility of reverse causation, all analyses were performed with a 1-year time lag.
The association between serum lipid levels and the risk of PD was evaluated using a time-dependent Cox proportional hazard regression, and the effect size was estimated as the HR. In addition, the crude incidence rate (number of PD cases per 100,000 person-years) for each serum lipid level was calculated. The middle tertile for each serum lipid level was used as the reference value. The index date was the date of the first blood test, and any changes in the lipid tertiles over time were updated as separate time segments. The participants were followed up until PD development, death, or December 31, 2019. Among individuals who developed PD during follow-up, we performed preliminary study to evaluate whether serum lipid levels over time are related to PD complications, such as dementia in PD (ICD-10 code F023) or death due to PD. Statistical significance was defined as a two-sided p value < 0.05. All analyses were performed using the SAS Enterprise Guide software (version 7.1; SAS Institute, Cary, NC, USA).
The datasets generated and/or analyzed during the current study are not publicly available due because the NHIS exclusively allows authorized persons to access data in a separate space. Upon an individual researcher's data set request, NHIS provides customized data to the researcher from the National Health Insurance Sharing Service at https://nhiss.nhis.or.kr/bd/ab/bdabb006iv.do.
Ascherio, A. & Schwarzschild, M. A. The epidemiology of Parkinson’s disease: Risk factors and prevention. Lancet Neurol. 15, 1257–1272 (2016).
Poewe, W. et al. Parkinson disease. Nat. Rev. Dis. Primers 3, 1–21 (2017).
Balestrino, R. & Schapira, A. Parkinson disease. Eur. J. Neurol. 27, 27–42 (2020).
Hu, G., Antikainen, R., Jousilahti, P., Kivipelto, M. & Tuomilehto, J. Total cholesterol and the risk of Parkinson disease. Neurology 70, 1972–1979 (2008).
Nam, G. E. et al. Metabolic syndrome and risk of Parkinson disease: A nationwide cohort study. PLos Med. 15, e1002640 (2018).
Simon, K. C., Chen, H., Schwarzschild, M. & Ascherio, A. Hypertension, hypercholesterolemia, diabetes, and risk of Parkinson disease. Neurology 69, 1688–1695 (2007).
Benn, M., Nordestgaard, B. G., Frikke-Schmidt, R. & Tybjærg-Hansen, A. Low LDL cholesterol, PCSK9 and HMGCR genetic variation, and risk of Alzheimer’s disease and Parkinson’s disease: Mendelian randomisation study. BMJ 357, j3170 (2017).
Fang, F. et al. Lipids, Apolipoproteins, and the risk of Parkinson disease: a prospective cohort study and a Mendelian randomization analysis. Circ. Res. 125, 643–652 (2019).
Fu, X. et al. A systematic review and meta-analysis of serum cholesterol and triglyceride levels in patients with Parkinson’s disease. Lipids Health Dis. 19, 1–10 (2020).
Huang, X., Abbott, R. D., Petrovitch, H., Mailman, R. B. & Ross, G. W. Low LDL cholesterol and increased risk of Parkinson’s disease: prospective results from Honolulu-Asia Aging Study. Mov. Disord. 23, 1013–1018 (2008).
Huang, X. et al. Statins, plasma cholesterol, and risk of Parkinson’s disease: a prospective study. Mov. Disord. 30, 552–559 (2015).
Park, J.-H. et al. Association of high-density lipoprotein cholesterol variability and the risk of developing Parkinson disease. Neurology 96, e1391–e1401 (2021).
Rozani, V. et al. Higher serum cholesterol and decreased Parkinson’s disease risk: a statin-free cohort study. Mov. Disord. 33, 1298–1305 (2018).
Sääksjärvi, K., Knekt, P., Männistö, S., Lyytinen, J. & Heliövaara, M. Prospective study on the components of metabolic syndrome and the incidence of Parkinson’s disease. Parkinsonism Relat. Disord. 21, 1148–1155 (2015).
Powers, K. M. et al. Dietary fats, cholesterol and iron as risk factors for Parkinson’s disease. Parkinsonism Relat. Disord. 15, 47–52 (2009).
Tan, L. C. et al. Dietary cholesterol, fats and risk of Parkinson’s disease in the Singapore Chinese Health Study. J. Neurol. Neurosurg. Psychiatry 87, 86–92 (2016).
Huang, X. et al. Serum cholesterol and the progression of Parkinson’s disease: results from DATATOP. PLoS ONE 6, e22854 (2011).
Björkhem, I. & Meaney, S. Brain cholesterol: Long secret life behind a barrier. Arterioscler. Thromb. Vasc. Biol. 24, 806–815 (2004).
Pfrieger, F. W. Role of cholesterol in synapse formation and function. Biochim. Biophys. Acta - Biomembr. 1610, 271–280 (2003).
Liu, Q. et al. Neuronal LRP1 knockout in adult mice leads to impaired brain lipid metabolism and progressive, age-dependent synapse loss and neurodegeneration. J. Neurosci. 30, 17068–17078 (2010).
Dietschy, J. M. & Turley, S. D. Thematic review series: Brain lipids. Cholesterol metabolism in the central nervous system during early development and in the mature animal. J. Lipid Res. 45, 1375–1397 (2004).
Saher, G. et al. High cholesterol level is essential for myelin membrane growth. Nat. Neurosci. 8, 468–475 (2005).
Dupuis, L. et al. Dyslipidemia is a protective factor in amyotrophic lateral sclerosis. Neurology 70, 1004–1009 (2008).
Block, R. C., Dorsey, E. R., Beck, C. A., Brenna, J. T. & Shoulson, I. Altered cholesterol and fatty acid metabolism in Huntington disease. J. Clin. Lipidol. 4, 17–23 (2010).
Karasinska, J. M. & Hayden, M. R. Cholesterol metabolism in Huntington disease. Nat. Rev. Neurol. 7, 561–572 (2011).
Wingo, T. S. et al. Association of early-onset Alzheimer disease with elevated low-density lipoprotein cholesterol levels and rare genetic coding variants of APOB. JAMA Neurol. 76, 809–817 (2019).
García-Sanz, P., Aerts, J. M. F. G. & Moratalla, R. The role of cholesterol in α-synuclein and lewy body pathology in GBA1 parkinson’s disease. Mov. Disord. 36, 1070–1085 (2021).
Jin, U., Park, S. J. & Park, S. M. Cholesterol metabolism in the brain and its association with Parkinson’s disease. Exp. Neurobiol. 28, 554 (2019).
Klemann, C. J. et al. Integrated molecular landscape of Parkinson’s disease. NPJ Parkinson’s Dis. 3, 1–7 (2017).
Hu, L. et al. Integrated metabolomics and proteomics analysis reveals plasma lipid metabolic disturbance in patients with Parkinson’s disease. Front. Mol. Neurosci. 13, 80 (2020).
Xicoy, H. et al. Shared genetic etiology between Parkinson’s disease and blood levels of specific lipids. NPJ Parkinson’s dis. 7, 1–8 (2021).
Seong, S. C. et al. Cohort profile: the national health insurance service-national health screening cohort (NHIS-HEALS) in Korea. BMJ Open 7, e016640 (2017).
CheolSeong, S. et al. Data resource profile: the national health information database of the National Health Insurance Service in South Korea. Int. J. Epidemiol. 46, 799–800 (2017).
This research was supported by a grant from the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, South Korea (Grant number: HI20C1130). Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.
The authors declare no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Hurh, K., Park, M., Jang, Si. et al. Association between serum lipid levels over time and risk of Parkinson’s disease. Sci Rep 12, 21020 (2022). https://doi.org/10.1038/s41598-022-25180-8
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.