Exploring the direct or mediating associations between lipids, atherosclerosis, obesity, and intervertebral disc degeneration: a Mendelian randomization study

doi:10.21203/rs.3.rs-3221400/v1

Download PDF

Research Article

Exploring the direct or mediating associations between lipids, atherosclerosis, obesity, and intervertebral disc degeneration: a Mendelian randomization study

https://doi.org/10.21203/rs.3.rs-3221400/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Objective

The aim of this study was to investigate whether there is a causal relationship between different lipids and intervertebral disc degeneration, and to verify whether obesity and atherosclerosis mediate the relationship between lipids and intervertebral disc degeneration.

Methods

Instrumental variables and their associated data were extracted from the corresponding genome-wide association studies. Inverse variance weighted (IVW), Weighted Median Estimator (WME), MR-Egger regression method, Simple mode (SM) method, Leave-One-Out method and Weighted Mode (WM) method were used to perform Mendelian randomization (MR) statistical analysis of the instrumental variables of exposure and outcome, Sensitivity analyses were then performed using Cochrane's Q value and MR − Egger intercept.

Results

In the two-sample MR Analysis, HDL, ApoA-1 were protective for both cervical disc degeneration (CDD) and non-cervical disc degeneration (NCDD), BMI contributed to both CDD and NCDD, while AS only contributed to NCDD. In the multivariate MR Analysis, the protective effects of HDL and ApoA-1 on CDD and NCDD were not independent of each other, Meanwhile, AS and BMI were independent risk factors for NCDD. In mediating MR Analysis, both AS and BMI mediated the effect of HDL on NCDD (AS 32.58%, BMI 15.49%), both AS and BMI mediated the effect of ApoA-1 on NCDD (AS 19.35%, BMI 17.65%).

Conclusion

HDL, ApoA-1, BMI are causally associated with both CDD and NCDD. The protective effects of HDL and ApoA-1 on intervertebral disc degeneration were not independent of each other. AS is causally associated with NCDD. BMI and AS are independent association factors for NCDD. BMI and AS also play important mediating roles in NCDD.

Intervertebral disc degeneration

Blood lipids

Atherosclerosis

Obesity

BMI

GWAS

SNP

Mendelian randomization

Intervertebral disc is a soft tissue structure with hydrodynamic characteristics between adjacent vertebral bodies. It is composed of the central nucleus pulposus, the peripheral annulus fibrosus, and the upper and lower covered cartilage endplate(1). Intervertebral disc degeneration (IVDD) is an incurable degenerative disease, which is mainly characterized by dehydration of the intervertebral disc, loss of intervertebral height, and cartilage endplate calcification(2), It is the pathological basis of discogenic neck pain, cervical spondylosis, non-specific low back pain and other common clinical diseases(3). Several studies have shown that lipid metabolism is closely related to IVDD(4). A retrospective case study including 302 patients showed that high-density lipoprotein (HDL) and triglyceride (TG) are important factors affecting the degree of intervertebral disc degeneration(5). An observational study found that high LDL levels was strongly associated with L4/5 disc degeneration (OR, 2.65) (6). In addition to the close relationship between lipoproteins and IVDD, an in vitro study has shown that leptin can initiate intervertebral disc degeneration, which is enhanced in an inflammatory environment(7). A GWAS study involving 303 patients also showed that leptin is closely related to IVDD, and leptin gene variants may affect the pathogenesis pathway of IVDD(8). Apolipoproteins have also been shown to be associated with disc degeneration(9). Therefore, this study will evaluate the causal relationship between lipid metabolism and intervertebral disc degeneration from six aspects: high density lipoprotein (HDL), low density lipoprotein (LDL), triglyceride (TG), leptin, apolipoprotein A-1 (ApoA-1) and apolipoprotein B (ApoB).

Previous studies have suggested that decreased blood supply and increased mechanical stress are important causes of IVDD, which are caused by atherosclerosis (AS) and obesity, respectively(10–12). Coincidentally, both AS and obesity are closely related to abnormal lipid metabolism. A meta-analysis that included more than 20,000 patients clearly showed a positive association between LDL levels and AS(13). Menno et al. demonstrated that HDL has an anti-AS effect(14). A meta-analysis of 43,068 subjects showed inverse associations between HDL levels, ApoA-I levels, and atherosclerotic vascular disease (ASCAD) (15). A Mendelian randomization study demonstrated that high TG could increase the risk of AS(16). Several studies have demonstrated that leptin exerts various pro-atherosclerosis effects. In addition, altered HDL distribution patterns and abnormal HDL metabolism can also be observed in obese individuals(17), Obesity is also associated with increased plasma LDL and oxLDL levels(18). A cohort study including 286 subjects demonstrated a significant association between obesity and TG(19). However, there is no direct evidence that dyslipidemia leads to IVDD by inducing AS or obesity. Therefore, we hope to verify the mediating effect of AS and obesity on the relationship between dyslipidemia and IVDD.

This study applied Mendelian randomization (MR) to explore the causal relationship between factors related to lipid metabolism, including HDL, LDL, TG, ApoA-I, ApoB, leptin, and IVDD, and to evaluate the mediating effect of AS and obesity between them. In recent years, MR is a new data analysis method that uses genetic variation as instrumental variable for epidemiological or etiological inference. It was first proposed by Professor Katan in 1986(20). Traditional observational epidemiological studies have large biases in identifying disease etiology and inferring causality, such as reverse causality, potential confounders, and exposure factors with small effects. The exposure and outcome factors of MR Are analogous to randomized controlled trials randomly assigned to treatment group and control group. The genetic variation will not be changed by other factors and people are formed at birth, which can effectively reduce the influence of confounding factors and reverse causality on the analysis(21). In this study, we used a two-sample, multivariate and mediated MR approach to explore the association of factors related to lipid metabolism with IVDD, In addition, we further explored the mediating effect of AS and obesity on lipid metabolism and IVDD.

Genome-wide association study summary statistics

A) Summary data on lipids and apolipoproteins in this study were obtained from the MRC Comprehensive Epidemiology Unit and included TG, HDL, LDL, ApoA-1, ApoB, summarized by Tom G. Richardson in 2020 (https://gwas.mrcieu.ac.uk/datasets/ieu-b-(107—111)/). It included 403,943 European populations from the UK Biobank, of which 12,321,875 SNPs were associated with lipids and apolipoproteins(22).

B) Genomic data on circulating leptin concentrations in this study were obtained from the UK Biobank project. In 2020, Yaghootkar et al. identified 203,423 SNPs associated with circulating leptin concentrations in an exome-based genotyping array study of 49,909 European subjects. They also determined that the two independent variants most strongly associated with circulating leptin concentrations were located in or near the LEP gene(23).

C) GWAS data of atherosclerosis in this study were obtained from FinnGen database (https://r9.finngen.fi/pheno/I9_ATHSCLE). FinnGen is a database of large public-private partnerships. This database collects and analyzes genomic and health data from 500,000 Finnish biobank participants. The study included 213,140 participants, 6599 with atherosclerosis and 206,541 controls. 16,380,428 SNPs were observed to be associated with atherosclerosis, and the study excluded the influence of cerebral arteries, coronary atherosclerosis, and peripheral artery disease.

D) Most current epidemiological data on obesity are based on BMI⁽²⁴⁾. The GWAS data of BMI in this study were obtained from the genome-wide association study (GWAS) of human BMI by Hoffmann TJ et al. The study included 234,069 individuals of European ancestry from the GIANT Consortium and 81,278 non-Hispanic whites from the Adult Genes, Environment, and Health Research Initiative (RPGEH) GERA cohort, involving 27,854,527 associated SNPs. Finally, 30 new independent BMI loci (P < 5.0 × 10⁻⁸) replicates were identified(25).

E) In this study, GWAS data related to IVDD phenotypes were divided into cervical disc degeneration (CDD) and non-cervical disc degeneration (NCDD), The above data are from the FinnGen database. CDD included 7,247 cases and 164,682 controls , 16,380,237 SNPs were associated with CDD (https://r9.finngen.fi/pheno/M13_CERVICDISC). NCDD included 20,001 cases and 164,682 controls, 16,380,337 SNPs were associated with NCDD (https://r9.finngen.fi/pheno/M13_INTERVERTEB).

Mendelian randomization statistical analysis

In this study, two-sample Mendelian randomization was first used to verify the causal relationship between HDL and other lipid metabolism-related factors and IVDD. After that, The risk factors with causal relationship were subjected to multivariate MR Analysis with AS and BMI respectively, This was used to verify whether AS or BMI were independent risk factors. Next, Mediating MR Was used to further analyze the mediating effect of AS and BMI and explore their mediating effect on other risk factors. The mediating effect mediated by AS or BMI was obtained by β₁ × β₂, β₁ is the MR Effect of other risk factors on AS and BMI, while β₂ is the MR Effect of AS and BMI on IVDD adjusted for other risk factors, β₃ is the MR Effect of other risk factors on IVDD after adjustment for AS and BMI. We estimated the proportion of mediating effects using the following formula (Standard errors were estimated using the Delta method(26)) :

$$E\left(\%\right)=\frac{{\sum }_{k=1}^{k}{\beta }_{1}\times {\beta }_{2}k}{{\sum }_{k=1}^{k}{\beta }_{3}+{{\beta }_{1}\times \beta }_{2}k}$$

Mediating effect calculation formula

$$SE\left({\beta }_{1}{\beta }_{2}\right)=\sqrt{{{\beta }_{1}}^{2}{SE\left({\beta }_{2}\right)}^{2}+{{\beta }_{2}}^{2}{SE\left({\beta }_{1}\right)}^{2}}$$

$$95\%CI\left({\beta }_{1}{\beta }_{2}\right)=\left({\beta }_{1}{\beta }_{2}\right)\pm 1.96 SE\left({\beta }_{1}{\beta }_{2}\right)$$

Delta method standard error estimation formula

Two-sample MR Analysis selected common SNPs with minimum allele frequency > 1% from pooled data for each exposure, Loci that remain horizontally significant throughout the genome (P ≤ 5×10^− 8), SNPs selected as IVs with linkage disequilibrium (r² = 0.001 and KB = 10000) and a minor allele frequency of less than 0.01 were removed, The independence of instrumental variables was ensured by excluding the effect of link-age disequilibrium (LD) on the results. At the same time, the effect alleles of single nucleotide polymorphisms (SNPs) were aligned, and all SNP with palindromic structure were removed to obtain significant SNP(27). We further excluded those loci that could not be reconciled with the corresponding SNP information from the outcome GWAS, Loci with a greater association with outcome than exposure were removed by using MR Steiger filtering(28). The proportion of exposure variation (R²) that could be explained by instrumental variables was calculated :R² = 2 × β²× (1-EAF) × EAF / (SE ×$\sqrt{\varvec{N}})$² (EAF, effect allele frequency; SE, Standard Error; N, number of cases; β, effect of each SNP on the exposure)(29). Next, weak instrumental variable bias of SNPs was evaluated by calculating the F-statistic: F = (N – k – 1) / k × R²/ (1 – R²) (N, number of cases; k, number of SNPs) (30), If F > 10, there is little possibility of introducing weak instrumental variable bias.

Inverse variance weighted (IVW) was used as the main MR Analysis method. The Beta value obtained from the results was converted into Odds Ratio (OR), and the 95% Confidence Interval (CI) was calculated. The Weighted Median Estimator (WME), MR-Egger regression, Simple mode (SM), Leave-One-Out and Weighted Mode (WM) as an auxiliary sensitivity analysis method. IVW assumes that all instrumental variables are valid instrumental variables, does not consider the intercept term and uses the inverse of the outcome variance as the weight for model fitting(31). In WM, SNPs with similar causal effects were clustered using a mode-based estimation model, and the causal effect estimation was obtained by weighting the impact of each SNP on the clustering by the inverse variance of its outcome effect. The estimation result was still robust even in the presence of a small number of invalid instrumental variables. WME and MR-Egger can provide valid causal estimates if at least half of the instrumental variables used are valid(32, 33), In addition, MR-Egger can evaluate the level pleiotropy of instrumental variables and provide a pleiotropy correction estimate, indicating the presence of pleiotropy when the regression intercept term is not zero and P < 0.05. Heterogeneity test was performed by Cochran's Q test to assess the heterogeneity of causal estimates for each SNP, and P < 0.05 was considered to indicate heterogeneity(34). Leave-one-out is a method to assess whether the MR Results are sensitive to this instrumental variable by removing individual relevant SNP one by one and calculating the combined effect of the remaining SNPS. When there was only heterogeneity but no pleiotropy, the IVW random effects model approach was accepted. When horizontal pleiotropy was present, the MR-Egger results were preferentially accepted(35).

All MR Analyses were performed with the use of the "TwoSampleMR", "MVMR" and "MendelianRandomization" packages in R software (version 4.3.1).

The number of SNPs as instrumental variables for the six exposure factors associated with lipid metabolism varied from 3 to 299 (HDL: 290, LDL: 140, TG: 249, ApoA-1: 299, ApoB: 149, Leptin: 3), The number of SNPs for the instrumental variables of BMI and AS was 138 and 27, respectively. Sample size, ethnicity, literature source, and GWAS ID of the selected instrumental variables are summarized in Supplemental Table S1. The genome-wide significant SNPs extracted as instrumental variables were summarized in the Supplementary Table S2-9, and SNPs with F statistic < 10 were excluded from each instrumental variable

In univariate MR Analysis (Fig. 1, 2), disc degeneration was classified into 2 categories: cervical disc degeneration (CDD) and non-cervical disc degeneration (NCDD). Data from the primary method IVW showed that HDL and ApoA-1 were protective against CDD (HDL: OR 0.87, 95%CI 0.79–0.96, P = 3.40E^− 03;APOA-1: OR 0.86, 95%CI 0.78–0.95, P = 3.49E^− 03). BMI can promote CDD (BMI: OR 1.28, 95%CI 1.06–1.53, P = 9.60E^− 03). LDL, ApoB, TG, leptin and AS were not significantly associated with CDD (LDL: OR 0.92, 95%CI 0.82–1.03, P = 1.52E^− 01༛TG: OR 1.04, 95%CI 0.94–1.16, P = 3.99E^− 01༛ApoB: OR 0.93, 95%CI 0.84–1.02, P = 1.09E^− 01༛Leptin: OR 1.30, 95%CI 0.98–1.74, P = 7.33E^− 02; AS: OR 1.04, 95%CI 0.96–1.13, P = 3.01E^− 01), No horizontal pleiotropy was observed in sensitivity analyses. HDL, TG, ApoA-1, AS and BMI have protective effects on NCDD (HDL: OR 0.93, 95%CI 0.87–0.99, P = 2.16E^− 02༛APOA-1: OR 0.93, 95%CI 0.87–0.99, P = 3.00E^− 02༛TG: OR 1.08, 95%CI 1.01–1.15, P = 1.98E^− 02; AS: OR 1.07, 95%CI 1.02–1.12, P = 8.26E^− 03; BMI: OR 1.26, 95%CI 1.12–1.42, P = 1.59E^− 04). LDL, ApoB, and leptin were not significantly associated with NCDD (LDL: OR 0.99, 95%CI 0.92–1.07, P = 8.44E^− 01༛ApoB: OR 1.03, 95%CI 0.97–1.10, P = 3.26E^− 01༛Leptin: OR 1.07, 95%CI 0.77–1.36, P = 8.68E^− 01). The results of MR-Egger were preferentially accepted because horizontal pleiotropy of IVs of TG was observed in the sensitivity analysis (TG: OR 0.96, 95%CI 0.87–1.06, P = 4.58E^− 01), Therefore, we concluded that TG was not significantly associated with NCDD.

The IVW, WME, MR-Egger, WM and MR-PRESSO results of CDD and NCDD, as well as the sensitivity analysis results of Cochran's Q and MR-Egger intercept are shown in Supplementary Table S10-11.

The results of two-sample MR Showed that BMI, AS, ApoA-1 and HDL were closely related to NCDD, Meanwhile, BMI, HDL and ApoA-1 were closely related to CDD. To verify whether BMI or AS played a mediating role in the relationship between HDL, ApoA-1 and disc degeneration, we first examined the causal relationship between HDL, ApoA-1, BMI and AS and NCDD by multivariable Mendelian randomization (MVMR), The results of MVMR showed that after adjusting for the effects of other exposures, HDL and ApoA-1 were no longer closely associated with NCDD (HDL: OR 0.87, 95%CI 0.70–1.09, P = 2.30E^− 01; ApoA-1: OR 1.09, 95%CI 0.87–1.37, P = 4.66E^− 01), instead, BMI and AS were still strongly associated with the incidence of NCDD (BMI: OR 1.23, 95%CI 1.09–1.39, P = 8.43E-04; AS: OR 1.07, 95%CI 1.02–1.13, P = 1.05E-02), indicating that BMI and AS were independent risk factors for NCDD. Next, we examined the causal relationship between HDL, ApoA-1, BMI and CDD by MVMR, The results of MVMR showed that HDL, ApoA-1 and BMI were no longer closely related to CDD after correcting the interaction between the three factors (BMI: OR 1.12, 95%CI 0.93–1.35, P = 2.24E-01; HDL: OR 1.19, 95%CI 0.86–1.66, P = 2.92E-01; ApoA-1: OR 0.73, 95%CI 0.52–1.03, P = 7.61E-02). In conclusion, BMI and AS mediate the relationship between HDL, ApoA-1 and NCDD.

Mediation Mendelian randomization was used to assess the mediating effect of non-independent risk factors on outcomes through other independent risk factors, The mediating effect of BMI and AS on the relationship between HDL, ApoA-1 and NCDD was further verified by mediating Mendelian randomization (Fig. 4). Our analysis showed that the BMI-adjusted OR for HDL on NCDD decreased from 0.92 (95%CI 0.87–0.98) to 0.94 (95%CI 0.87–1.01), The mediating effect of BMI on the relationship between HDL and NCDD was 15.49% (95%CI 3.23%-27.71%). The effect of ApoA-1 on NCDD decreased from 0.93(95%CI 0.87–0.99) to 0.94 (95%CI 0.87-1.00) after adjustment for BMI, BMI accounted for 17.65% (95%CI 2.78%-32.03%) of the mediating risk factors between ApoA-1 and NCDD. The effect of HDL on NCDD decreased from 0.92 (95%CI 0.87–0.98) to 0.95 (95%CI 0.89–1.01) after adjustment for AS, AS accounted for 32.58% (95%CI 14.29%-46.43%) of the risk factors mediating between HDL and NCDD. The effect of ApoA-1 on NCDD decreased from 0.93(95%CI 0.87–0.99) to 0.94 (95%CI 0.88-1.00) after adjustment for AS, The proportion of AS in the mediating risk factors between ApoA-1 and NCDD was 19.35% (95%CI 1.64%-38.78%). Multivariate MR Analysis data are summarized in Supplementary Table S12-16. The calculation data of the proportion of intermediaries are summarized in the Supplementary Table S17-18.

In this study, we found that HDL and ApoA-1, which are related to lipid metabolism, are associated with CDD and NCDD, and they are protective against IVDD in different segments. however, the causal relationship between HDL and ApoA-1 and IVDD is not independent, We demonstrated that they were associated with NCDD mediated by BMI and AS, and there was no evidence that BMI and AS mediated the relationship between HDL and ApoA-1 and CDD.

Consistent with the results of our study, most studies have suggested that HDL is closely related to IVDD. A retrospective clinical study involving 302 patients found that the top two factors affecting the severity classification of IVDD were age (32.9%) and HDL (20.7%), and suggested that increased HDL levels may reduce the incidence of IVDD⁽⁵⁾. A similar conclusion was reached by another retrospective study including 790 patients, who found that the ratio of TC/HDL-C and LDL-C/HDL-C were significantly associated with lumbar disc herniation (LDH) (36). Recently, a comprehensive MR Study demonstrated 13 blood metabolites significantly associated with IVDD among 249 blood metabolites, 12 of which were HDL-related metabolites(37), This further confirms the close relationship between HDL and IVDD. There is no clear evidence that there is a causal relationship between ApoA-1 and IVDD. As one of the main components of HDL, ApoA-1 is responsible for interacting with lecithin/cholesterol acyltransferase (LCAT) and other intracellular receptors(38), It’s level and structural stability determine the rate of HDL formation(39). Therefore, in this study, we identified a causal relationship between ApoA-1 and IVDD in different segments and demonstrated that HDL did not have a protective effect on IVDD independent of ApoA-1.

The close relationship between HDL and ApoA-1 and BMI has been demonstrated in several studies, and a cross-sectional study involving 333 men found an inverse association between HDL and BMI(40). A recent cross-sectional study involving 348 patients reached the same conclusion(41). A meta-analysis study also pointed out that the increase of BMI was directly related to the decrease of ApoA-1 level(42). Another cohort study including 1051 subjects found an inverse association between waist-to-height ratio and ApoA-1 levels(43). Meanwhile, BMI was significantly associated with the occurrence of IVDD, Because the causal relationship between HDL and ApoA-1 and IVDD is not independent, therefore, we analyzed the mediating effect of BMI between them in this study. We found that HDL and ApoA-1 were associated with NCDD through BMI as a mediator, and the mediating effect of BMI on ApoA-1 was higher than that of HDL. In addition, we found that BMI did not mediate the relationship between HDL or ApoA-1 and CDD, which may imply that non-cervical discs are more susceptible to mechanical degeneration than cervical discs, A recent cross-sectional study also found that the degeneration of cervical discs, unlike lumbar discs, was not related to the support and restraint capacity of paravertebral muscles(44).

HDL can transport extrahepatic lipid waste into the liver for metabolism, and finally be excreted through bile, and its anti-AS function has been recognized. However, with the failure of Torcetrapib (a selective CETP inhibitor) in clinical trials, people doubted the anti-AS function of HDL(45), The latest research direction has gradually shifted from improving the absolute value of blood HDL-C to improving the reverse cholesterol transport (RCT) function of HDL(46). Current studies suggest that the adaptation of ApoA-1 conformation is a determinant of HDL function(47), Compared with cholesteryl ester transfer protein (CETP) and other drugs, peptide mimetics of ApoA-1 have a better and faster improvement effect on low HDL syndrome(48). In addition, several earlier observational studies have demonstrated a significant interaction between IVDD and AS(49, 50), It is considered that disc "ischemia" caused by AS is an important cause of the occurrence and development of IVDD, A meta-analysis also pointed out that AS was associated with IVDD(51), However, a recent meta-analysis concluded that there was insufficient evidence to prove the association between ischemia and IVDD(52). Meanwhile, previous studies only focused on the interaction between lumbar disc degeneration and AS, Therefore, we explored the causal relationship between AS and IVDD in different segments by MR Method and demonstrated that there was no causal relationship between AS and CDD, while AS promoted NCDD. In this study, we also found that AS mediated the relationship between HDL and ApoA-1 and NCDD, and the effect of AS was stronger than that of BMI.

Inconsistently, we only found GWAS data on atherosclerosis excluding heart and brain and IVDD data with cervical/non-cervical classification in the FinnGen database, As a result, there will be a partial overlap between the two samples, which may cause certain limitations in the final result. In addition, the percentage of mediation effects may be biased by the nonoverlapping nature of odds ratios (OR). Finally, only people of European ancestry were included in this study, and the relevant results of this study are only applicable to people of European ancestry, and other ancestral groups need to be further verified.

Both HDL and ApoA-1 have a causal relationship with CDD/NCDD, and both HDL and ApoA-1 have a protective effect on IVDD, which is not independent of each other. In addition, there was a causal relationship between AS and NCDD, but no causal relationship was found between AS and CDD. Finally, BMI and AS only mediated NCDD, AS played a major mediating role between HDL and NCDD, and the mediating role of AS between ApoA-1 and NCDD was similar to that of BMI.

ApoA-1 Apolipoprotein A-I

ApoB Apolipoprotein B

BMI Body Mass Index

CDD Cervical disc degeneration

CETP Cholesteryl ester transfer protein

GWAS Genome-wide association study

HDL High-density lipoprotein

IVDD Intervertebral disc degeneration

IVW Inverse variance weighted

LDL Low Density Lipoprotein

MR Mendelian randomization

NCDD Non-cervical disc degeneration

OR Odds Ratio

SNP Single nucleotide polymorphism

TG Triglyceride

WM Weighted Mode

WME Weighted Median Estimator

1.Ethics approval and consent to participate

All analyses were based on publicly available summary statistics, which do not require ethical approval and consent.

2. Consent for publication

Not applicable.

3. Availability of data and materials

GWAS statistics of atherosclerosis and Intervertebral disc degeneration were obtained from FinnGen database:

(https://r9.finngen.fi/pheno/I9_ATHSCLE,https://r9.finngen.fi/pheno/M13_CERVICDISC,https://r9.finngen.fi/pheno/M13_INTERVERTEB).

GWAS statistics of high density lipoprotein, low density lipoprotein, triglyceride, apolipoprotein A-1, apolipoprotein B, leptin and Body mass index were obtained from MRC-IEU OpenGWAS:

(https://gwas.mrcieu.ac.uk/datasets/ieu-b-107/,https://gwas.mrcieu.ac.uk/datasets/ieu-b-108/,https://gwas.mrcieu.ac.uk/datasets/ieu-b-109/,https://gwas.mrcieu.ac.uk/datasets/ieu-b-110/,https://gwas.mrcieu.ac.uk/datasets/ieu-b-111/,https://gwas.mrcieu.ac.uk/datasets/ebi-a-GCST90007310/, https://gwas.mrcieu.ac.uk/datasets/ebi-a-GCST006368/).

The GWAS summary statistics used in this study is available from supplemental instrument.

4. Competing interests

All authors declare no conflicts of interest.

5.Funding

None.

6. Authors’ contributions

YANG Yuhang, ZHANG Naijin and LIU Aifeng designed the study. YANG Yuhang, GUO Tianci and CHEN Jiaming wrote and edited the manuscript. ZHANG Naijin and YANG Zhenghui collected and analyzed the data. JIANG Xingyu drew the figures. YANG Yuhang and WANG Ping had full access to all the data in the study and had final responsibility for the decision to submit for publication. All authors read and approved the final manuscript.

7. Acknowledgements

Not applicable.

Desmoulin GT, Pradhan V, Milner TE. Mechanical Aspects of Intervertebral Disc Injury and Implications on Biomechanics. Spine (Phila Pa 1976) 2020;45(8):E457–E464.
Cao G, Yang S, Cao J, Tan Z, Wu L, Dong F, Ding W, Zhang F. The Role of Oxidative Stress in Intervertebral Disc Degeneration. Oxid Med Cell Longev 2022;2022:2166817.
Li B, Yang Y, Wang L, Liu G. Stem Cell Therapy and Exercise for Treatment of Intervertebral Disc Degeneration. Stem Cells Int 2021;2021:7982333.
Teraguchi M, Yoshimura N, Hashizume H, Muraki S, Yamada H, Oka H, Minamide A, Ishimoto Y, Nagata K, Kagotani R, Tanaka S, Kawaguchi H, Nakamura K, Akune T, Yoshida M. Metabolic Syndrome Components Are Associated with Intervertebral Disc Degeneration: The Wakayama Spine Study. PLoS One 2016;11(2):e0147565.
Huang Z, Chen J, Su Y, Guo M, Chen Y, Zhu Y, Nie G, Ke R, Chen H, Hu J. Impact of dyslipidemia on the severity of symptomatic lumbar spine degeneration: A retrospective clinical study. Front Nutr 2022;9:1033375.
Hangai M, Kaneoka K, Kuno S, Hinotsu S, Sakane M, Mamizuka N, Sakai S, Ochiai N. Factors associated with lumbar intervertebral disc degeneration in the elderly. Spine J 2008;8(5):732–740.
Segar AH, Fairbank JCT, Urban J. Leptin and the intervertebral disc: a biochemical link exists between obesity, intervertebral disc degeneration and low back pain-an in vitro study in a bovine model. Eur Spine J 2019;28(2):214–223.
Chen H-H, Hsu H-T, Liao M-H, Teng M-S. Effects of Sex and Obesity on LEP Variant and Leptin Level Associations in Intervertebral Disc Degeneration. Int J Mol Sci 2022;23(20):12275.
Zhou Y, Chen X, Tian Q, Zhang J, Wan M, Zhou X, Xu X, Cao X, Zhou X, Zheng L. Deletion of ApoE Leads to Intervertebral Disc Degeneration via Aberrant Activation of Adipokines. Spine (Phila Pa 1976) 2022;47(12):899–907.
Kasperczak M, Kuciel-Lewandowska J, Gnus J, Paprocka-Borowicz M. Assessment of Changes in Lipids Metabolism in Patients with Degenerative Joints and Discs Diseases Subjected to Spa Therapy. Biomed Res Int 2019;2019:4732654.
Liu Z, Fu C. Application of single and cooperative different delivery systems for the treatment of intervertebral disc degeneration. Front Bioeng Biotechnol 2022;10:1058251.
Wilson Zingg R, Kendall R. Obesity, Vascular Disease, and Lumbar Disk Degeneration: Associations of Comorbidities in Low Back Pain. PM R 2017;9(4):398–402.
Ference BA, Ginsberg HN, Graham I, Ray KK, Packard CJ, Bruckert E, Hegele RA, Krauss RM, Raal FJ, Schunkert H, Watts GF, Borén J, Fazio S, Horton JD, Masana L, Nicholls SJ, Nordestgaard BG, van de Sluis B, Taskinen M-R, Tokgözoğlu L, Landmesser U, Laufs U, Wiklund O, Stock JK, Chapman MJ, Catapano AL. Low-density lipoproteins cause atherosclerotic cardiovascular disease. 1. Evidence from genetic, epidemiologic, and clinical studies. A consensus statement from the European Atherosclerosis Society Consensus Panel. Eur Heart J 2017;38(32):2459–2472.
Hoekstra M, Van Berkel TJC. Functionality of High-Density Lipoprotein as Antiatherosclerotic Therapeutic Target. Arterioscler Thromb Vasc Biol 2016;36(11):e87–e94.
Ali KM, Wonnerth A, Huber K, Wojta J. Cardiovascular disease risk reduction by raising HDL cholesterol – current therapies and future opportunities. Br J Pharmacol 2012;167(6):1177–1194.
Johansen CT, Hegele RA. Using Mendelian randomization to determine causative factors in cardiovascular disease. J Intern Med 2013;273(1):44–47.
Stadler JT, Marsche G. Obesity-Related Changes in High-Density Lipoprotein Metabolism and Function. Int J Mol Sci 2020;21(23):8985.
de Freitas MCP, Fernandez DGE, Cohen D, Figueiredo-Neto AM, Maranhão RC, Damasceno NRT. Oxidized and electronegative low-density lipoprotein as potential biomarkers of cardiovascular risk in obese adolescents. Clinics (Sao Paulo) 2018;73:e189.
Dimitriadis DG, Mamplekou E, Dimitriadis PG, Komessidou V, Papakonstantinou A, Dimitriadis GD, Papageorgiou C. The Association Between Obesity and Hostility: The Mediating Role of Plasma Lipids. J Psychiatr Pract 2016;22(3):166–174.
Freuer D, Linseisen J, Meisinger C. Association Between Inflammatory Bowel Disease and Both Psoriasis and Psoriatic Arthritis: A Bidirectional 2-Sample Mendelian Randomization Study. JAMA Dermatol 2022;158(11):1262–1268.
Wu H, Yang J, Wang H, Li L. Mendelian randomization to explore the direct or mediating associations between socioeconomic status and lung cancer. Front Oncol 2023;13:1143059.
Richardson TG, Sanderson E, Palmer TM, Ala-Korpela M, Ference BA, Davey Smith G, Holmes MV. Evaluating the relationship between circulating lipoprotein lipids and apolipoproteins with risk of coronary heart disease: A multivariable Mendelian randomisation analysis. PLoS Med 2020;17(3):e1003062.
Yaghootkar H, Zhang Y, Spracklen CN, Karaderi T, Huang LO, Bradfield J, Schurmann C, Fine RS, Preuss MH, Kutalik Z, Wittemans LBL, Lu Y, Metz S, Willems SM, Li-Gao R, Grarup N, Wang S, Molnos S, Sandoval-Zárate AA, Nalls MA, Lange LA, Haesser J, Guo X, Lyytikäinen L-P, Feitosa MF, Sitlani CM, Venturini C, Mahajan A, Kacprowski T, Wang CA, Chasman DI, Amin N, Broer L, Robertson N, Young KL, Allison M, Auer PL, Blüher M, Borja JB, Bork-Jensen J, Carrasquilla GD, Christofidou P, Demirkan A, Doege CA, Garcia ME, Graff M, Guo K, Hakonarson H, Hong J, Chen Y-DI, Jackson R, Jakupović H, Jousilahti P, Justice AE, Kähönen M, Kizer JR, Kriebel J, LeDuc CA, Li J, Lind L, Luan J, Mackey DA, Mangino M, Männistö S, Carli JFM, Medina-Gomez C, Mook-Kanamori DO, Morris AP, Mutsert R de, Nauck M, Prokic I, Pennell CE, Pradhan AD, Psaty BM, Raitakari OT, Scott RA, Skaaby T, Strauch K, Taylor KD, Teumer A, Uitterlinden AG, Wu Y, Yao J, Walker M, North KE, Kovacs P, Ikram MA, Duijn CM van, Ridker PM, Lye S, Homuth G, Ingelsson E, Spector TD, McKnight B, Province MA, Lehtimäki T, Adair LS, Rotter JI, Reiner AP, Wilson JG, Harris TB, Ripatti S, Grallert H, Meigs JB, Salomaa V, Hansen T, Dijk KW van, Wareham NJ, Grant SFA, Langenberg C, Frayling TM, Lindgren CM, Mohlke KL, Leibel RL, Loos RJF, Kilpeläinen TO. Genetic Studies of Leptin Concentrations Implicate Leptin in the Regulation of Early Adiposity. Diabetes 2020;69(12):2806.
Caballero B. Humans against Obesity: Who Will Win? Adv Nutr 2019;10(Suppl 1):S4–S9.
Hoffmann TJ, Choquet H, Yin J, Banda Y, Kvale MN, Glymour M, Schaefer C, Risch N, Jorgenson E. A Large Multiethnic Genome-Wide Association Study of Adult Body Mass Index Identifies Novel Loci. Genetics 2018;210(2):499–515.
Tofighi D, MacKinnon DP. RMediation: an R package for mediation analysis confidence intervals. Behav Res Methods 2011;43(3):692–700.
Hemani G, Zheng J, Elsworth B, Wade KH, Haberland V, Baird D, Laurin C, Burgess S, Bowden J, Langdon R, Tan VY, Yarmolinsky J, Shihab HA, Timpson NJ, Evans DM, Relton C, Martin RM, Davey Smith G, Gaunt TR, Haycock PC. The MR-Base platform supports systematic causal inference across the human phenome. Elife 2018;7:e34408.
Hemani G, Tilling K, Davey Smith G. Orienting the causal relationship between imprecisely measured traits using GWAS summary data. PLoS Genet 2017;13(11):e1007081.
Pang L, Ding Z, Chai H, Li F, Wu M, Shuang W. Causal relationship between smoking status, smoking frequency and bladder cancer: a Mendelian randomization study. Genes Genomics 2023;45(2):203–213.
Huang L, Xie Y, Jin T, Wang M, Zeng Z, Zhang L, He W, Mai Y, Lu J, Cen H. Diet-derived circulating antioxidants and risk of knee osteoarthritis, hip osteoarthritis and rheumatoid arthritis: a two-sample Mendelian randomization study. Front Med (Lausanne) 2023;10:1147365.
Burgess S, Thompson SG. Erratum to: Interpreting findings from Mendelian randomization using the MR-Egger method. European Journal of Epidemiology 2017;32(5):391.
Otomo N, Khanshour AM, Koido M, Takeda K, Momozawa Y, Kubo M, Kamatani Y, Herring JA, Ogura Y, Takahashi Y, Minami S, Uno K, Kawakami N, Ito M, Sato T, Watanabe K, Kaito T, Yanagida H, Taneichi H, Harimaya K, Taniguchi Y, Shigematsu H, Iida T, Demura S, Sugawara R, Fujita N, Yagi M, Okada E, Hosogane N, Kono K, Nakamura M, Chiba K, Kotani T, Sakuma T, Akazawa T, Suzuki T, Nishida K, Kakutani K, Tsuji T, Sudo H, Iwata A, Inami S, Wise CA, Kochi Y, Matsumoto M, Ikegawa S, Watanabe K, Terao C. Evidence of causality of low body mass index on risk of adolescent idiopathic scoliosis: a Mendelian randomization study. Front Endocrinol (Lausanne) 2023;14:1089414.
Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent Estimation in Mendelian Randomization with Some Invalid Instruments Using a Weighted Median Estimator. Genet Epidemiol 2016;40(4):304–314.
Carnegie R, Zheng J, Sallis H, Jones H, Wade K, Evans J, Zammit S, Munafò M, Martin R. Mendelian Randomization for Nutritional Psychiatry. Lancet Psychiatry 2020;7(2):208–216.
Burgess S, Thompson SG. Interpreting findings from Mendelian randomization using the MR-Egger method. Eur J Epidemiol 2017;32(5):377–389.
Zhang Y, Zhao Y, Wang M, Si M, Li J, Hou Y, Jia J, Nie L. Serum lipid levels are positively correlated with lumbar disc herniation—a retrospective study of 790 Chinese patients. Lipids Health Dis 2016;15:80.
Ji L, Wang Y, Lu T, Yang J, Luo C, Qiu B. Identification of blood metabolites linked to the risk of intervertebral disc diseases: a comprehensive Mendelian randomization study. Postgrad Med J 2023:qgad052.
Dergunov AD, Garaeva EA, Savushkin EV, Litvinov DY. Significance of Lipid-Free and Lipid-Associated ApoA-I in Cellular Cho-lesterol Efflux. Curr Protein Pept Sci 2017;18(1):92–99.
Vedhachalam C, Chetty PS, Nickel M, Dhanasekaran P, Lund-Katz S, Rothblat GH, Phillips MC. Influence of apolipoprotein (Apo) A-I structure on nascent high density lipoprotein (HDL) particle size distribution. J Biol Chem 2010;285(42):31965–31973.
Al-Ajlan AR. Lipid Profile in Relation to Anthropometric Measurements among College Male Students in Riyadh, Saudi Arabia: A Cross-Sectional Study. Int J Biomed Sci 2011;7(2):112–119.
Okekunle AP, Yie G-E, Song S, Kim Z, Youn HJ, Cho J, Min JW, Kim YS, Lee JE. Association of lipid profile with obesity among breast cancer survivors: a cross-sectional study. Lipids Health Dis 2022;21:66.
Forte F, Calcaterra I, Lupoli R, Orsini RC, Chiurazzi M, Tripaldella M, Iannuzzo G, Di Minno MND. Association of apolipoprotein levels with peripheral arterial disease: a meta-analysis of literature studies. Eur J Prev Cardiol 2022;28(18):1980–1990.
Wang W, Blackett P, Khan S, Lee E. Apolipoproteins A-I, B, and C-III and Obesity in Young Adult Cherokee. J Lipids 2017;2017:8236325.
Kumagai G, Wada K, Kudo H, Asari T, Chiba D, Ota S, Takeda O, Koyama K, Nakaji S, Ishibashi Y. Associations between cervical disc degeneration and muscle strength in a cross-sectional population-based study. PLoS ONE 2019;14(1). doi:10.1371/journal.pone.0210802.
Santos-Gallego CG, Ibanez B, Badimon JJ. HDL-cholesterol: is it really good? Differences between apoA-I and HDL. Biochem Pharmacol 2008;76(4):443–452.
Karalis I, Jukema JW. HDL Mimetics Infusion and Regression of Atherosclerosis: Is It Still Considered a Valid Therapeutic Option? Curr Cardiol Rep 2018;20(8):66.
Borja MS, Zhao L, Hammerson B, Tang C, Yang R, Carson N, Fernando G, Liu X, Budamagunta MS, Genest J, Shearer GC, Duclos F, Oda MN. HDL-apoA-I exchange: rapid detection and association with atherosclerosis. PLoS One 2013;8(8):e71541.
Pal M, Pillarisetti S. HDL elevators and mimetics--emerging therapies for atherosclerosis. Cardiovasc Hematol Agents Med Chem 2007;5(1):55–66.
Turgut AT, Sönmez I, Cakıt BD, Koşar P, Koşar U. Pineal gland calcification, lumbar intervertebral disc degeneration and abdominal aorta calcifying atherosclerosis correlate in low back pain subjects: A cross-sectional observational CT study. Pathophysiology 2008;15(1):31–39.
Kauppila LI, Penttilä A, Karhunen PJ, Lalu K, Hannikainen P. Lumbar disc degeneration and atherosclerosis of the abdominal aorta. Spine (Phila Pa 1976) 1994;19(8):923–929.
Kauppila LI. Atherosclerosis and disc degeneration/low-back pain--a systematic review. Eur J Vasc Endovasc Surg 2009;37(6):661–670.
Shcherbina A, Longacre M. The Association Between Atherosclerosis and Low Back Pain: A Systematic Review. PM R 2017;9(11):1144–1156.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Exploring the direct or mediating associations between lipids, atherosclerosis, obesity, and intervertebral disc degeneration: a Mendelian randomization study

Status:

Version 1

Abstract

Objective

Methods

Results

Conclusion

Figures

Introduction

Material and methods

Mendelian randomization statistical analysis

Mediating effect calculation formula

Delta method standard error estimation formula

Results

Discussion

Conclusion

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1