1. Introduction
Hyperlipidemia, characterized by elevated levels of total cholesterol, triglycerides, or low-density lipoprotein cholesterol, or reduced levels of high-density lipoprotein cholesterol, is a prevalent metabolic disorder and a major risk factor for atherosclerotic cardiovascular diseases, which remain the leading cause of morbidity and mortality worldwide.[1-3] The global prevalence of hyperlipidemia has been rising steadily, driven by changes in dietary patterns, sedentary lifestyles, and an aging population, imposing a substantial burden on healthcare systems.[4,5] Consequently, elucidating the underlying pathogenesis of hyperlipidemia and identifying novel intervention targets are of paramount importance for its prevention and management.
In recent years, the gut microbiota (GM), often described as the “second genome” of the human body, has emerged as a critical regulator of host metabolism and systemic health.[6,7] Residing in the human gastrointestinal tract, the GM comprises a complex ecosystem of bacteria, including predominant phyla such as Firmicutes and Bacteroidetes, which play essential roles in nutrient processing, immune modulation, and the production of bioactive metabolites.[8] Growing evidence suggests that disturbances in the composition and function of the GM – termed dysbiosis – are closely linked to the development of various metabolic disorders, including obesity, type 2 diabetes, and cardiovascular diseases.[9,10] In particular, a number of observational and experimental studies have indicated a potential association between specific gut microbial taxa and lipid metabolism. For instance, certain microbial-derived metabolites, such as short-chain fatty acids and bile acids, have been shown to influence host lipid homeostasis through mechanisms involving hepatic cholesterol synthesis, energy expenditure, and bile acid receptor signaling.[11,12] Moreover, interventional strategies targeting the GM, including probiotic supplementation and fecal microbiota transplantation, have demonstrated promising effects on improving lipid profiles in both animal models and clinical studies.[13-15] However, due to the inherent limitations of observational study designs – such as susceptibility to confounding factors (e.g., diet, medication use, and lifestyle) and the inability to establish the direction of causality – it remains unclear whether alterations in the GM are a cause or a consequence of hyperlipidemia.[16]
Mendelian randomization (MR) is a powerful epidemiological method that uses genetic variants as instrumental variables (IVs) to infer causal relationships, effectively mitigating the confounding and reverse causation biases inherent in observational studies.[17] Given its potential to address the causality dilemma outlined above, we employed a 2-sample bidirectional MR framework.[18] We leveraged summary-level data from large-scale genome-wide association studies (GWAS) to interrogate the causal link between gut microbial taxa and hyperlipidemia. Specifically, our study aimed to: assess whether specific gut microbial taxa exert a causal influence on hyperlipidemia risk, and examine the possibility of reverse causation. We anticipate that this investigation will offer novel insights into the etiological role of the GM and inform future microbiota-targeted strategies for managing hyperlipidemia.[10,19]
2. Materials and methods
2.1. Study design
This study used a 2-sample bidirectional MR analysis to evaluate the causal relationship between 196 intestinal microbiota and hyperlipidemia. The specific research design is illustrated in Figure 1. Among them, the intestinal microbiota is exposed, leading to hyperlipidemia.
The following 3 main MR presumptions must all be satisfied simultaneously to obtain reliable results: IVs are closely related to exposure; IVs have no relation to confounding factors; and IVs only affect the outcome through exposure.[20]
GM data were derived from the largest standardized meta-analysis of the genome-wide association study (GWAS) to date (http://www.mibiogen.org/), with a total sample size of 18,340 from 24 cohorts. There were 57,47,754 single nucleotide polymorphisms (SNPs).[8] Hyperlipidemia data were obtained from the UK Biobank via the IEU OpenGWAS project (https://gwas.mrcieu.ac.uk/) to access the number according to (GWAS ID: ebi-a-GCST90104007), which included 3,49,222 healthy European participants, with a total of 1,45,02,301 SNPs. Since the data used in this study are publicly available, no additional ethical approval or relevant consent was required.
The study included 196 intestinal microbiota samples after excluding 15 unknown bacterial genera from a total of 211 intestinal microbiota samples. IVs were screened using the following criteria: SNPs with a significance threshold (P < 5.0 × 10−5) within the genetic locus range were selected as potential IVs. The linkage disequilibrium threshold was set at r2 < 0.001; the aggregation window size was set at 10,000 kb to ensure that the included SNPs are independent of each other[18,21]; SNPs with F > 10 were selected to exclude weak IVs.
The following formula was used to calculate the F value: F = β2/selenium2, where β is the effect value and se is the standard error.[22]
2.2. MR analysis
The inverse variance weighted (IVW) method was adopted as the primary strategy. Its advantage lies in ignoring the presence of the intercept term in the regression and using the reciprocal of the result variance as the fitting weight. IVW yields the most reliable results when there is no heterogeneity and horizontal pleiotropy in the IVs.[23,24] Five methods were used for secondary confirmation, namely weighted mode,[25] MR-Egger regression,[26] simple model,[27] and weighted median.[28] The β value directions of the 5 methods were the same.
Finally, MR was used to determine the causal relationship between specific bacteria and 5 categories of hyperlipidemia using the same positive definition criteria.
2.3. Sensitivity analysis
Sensitivity analyses included tests for horizontal pleiotropy, heterogeneity, and elimination. The horizontal pleiotropy test was conducted through MR-Egger regression.[26] The study exhibits horizontal pleiotropy if the MR-Egger analysis has a significant intercept term. Cochran’s Q test was used to determine the heterogeneity of SNPs[24] A statistically significant (P ≤ .05) Cochran’s Q statistic test demonstrated that the analysis results had significant heterogeneity.
Mendelian randomization pleiotropy residual sum and outlier test (MR-PRESSO) was used to identify outliers.[29] If outliers were detected, they were removed, and the remaining IVs were reanalyzed. The retention method was used to determine whether a single SNP contributed to the significant results. The association between the human intestinal microbiota and the risk of lung cancer is expressed as an odds ratio (OR) and its 95% confidence interval (CI).
A P-value of ≤.05 indicates that there may be a causal relationship. Multiple testing can result in false positives when using sensitivity analysis and multiple regression models. The risk of false positives caused by multiplicity was effectively reduced by combining the preset analysis levels, false discovery rate correction, and the consistency requirements of the direction of sensitivity analyses.
Future studies will be used to further verify these findings through independent samples.
2.4. Statistical analysis
Data were analyzed using the 2-sample-MR (TSMR) and MR-PRESSO R packages of R software (version 4.3.1; R Core Team, R Foundation for Statistical Computing, https://www.r-project.org/). A difference with a P ≤ .05 was considered statistically significant.
3. Results
3.1. Results of screening for IVs
Finally, 9 different intestinal flora (2 orders, 2 families, and 5 genera) were selected for possible causal associations with hyperlipidemia based on the screening criteria for the study’s IVs. The 111 SNPs that were highly associated with hyperlipidemia were included as IVs, and the statistics F were all >10, indicating no weak IV bias (Table 1).
3.2. MR analysis
This study used 5 TSMR methods, with the IVW method serving as the main one. Finally, 9 intestinal flora were found to be associated with hyperlipidemia. Enterobacteriaceae (OR = 1.307, 95% CI: 1.054–1.621, P = .015), Adlercreutzia (OR = 1.254, 95% CI: 1.069–1.470, P = .005, LachnospiraceaeUCG001 (OR = 1.201, 95% CI: 1.004–1.446, P = .045), Ruminococcus1 (OR = 1.224, 95% CI: 1.010–1.483, P = .039), and Enterobacteriales (OR = 1.307, 95% CI: 1.054–1.621, P = .015) were associated with the onset of hyperlipidemia, increasing its risk, which was positively correlated; while Oxalobacteraceae (OR = 0.897, 95% CI: 0.804–1.000, P = .049), Clostridiumsensustricto1 (OR = 0.728, 95% CI: 0.594–0.894, P = .002), RuminococcaceaeUCG010 (OR = 0.792, 95% CI: 0.628–0.999, P = .049), and Rhodospirillales (OR = 0.846, 95% CI: 0.733–0.977, P = .022) may reduce the risk of developing hyperlipidemia and were negatively associated with its onset (Figs.
2–4; Table 2).
3.3. Sensitivity analysis
The horizontal pleiotropy between SNPs and hyperlipidemia results was confirmed using MR-Egger regression; all of which had P > .05, indicating that there was no pleiotropy in the analyses. MR-PRESSO was again applied to analyze the global test of the relevant microbiota communities (P > .05), demonstrating that there was no horizontal pleiotropy. Cochran’s test was used to assess heterogeneity, and the results revealed that IVW and MR-Egger’s Q P values for all the flora were >.05, indicating no heterogeneity between SNPs within each microbial taxonomic group (Table 3).
The leave-one-out test results confirmed the robustness of MR results, and no single SNP was found to bias the results significantly (Fig. 5).
3.4. Reverse MR analysis results
Reverse MR analyses were performed on the 9 gut flora identified in the forward MR analysis, with hyperlipidemia as the omission factor and gut flora as the outcome variable. The IVW analyses revealed that hyperlipidemia was not causally associated with the 9 distinct gut flora across classifications, and the tests for heterogeneity and multiplicity yielded negative results (Table 4). The results for each SNP associated with hyperlipidemia and 9 intestinal flora are displayed in Table 5.
4. Discussion
Hyperlipidemia is a common metabolic disorder and also a global health problem, as its incidence and prevalence rates are increasing worldwide. Sensitive biomarkers are recognized to be necessary for hyperlipidemia, and more importantly, specific biomarkers for hyperlipidemia will improve the clinical diagnosis and treatment in the early stages of the disease.[30,31] Mounting evidence suggests that GM can cause metabolic diseases,[32] making it a new research direction and a potential strategy for disease control. To date, no study has been able to fully elucidate the mechanism of action between GM and hyperlipidemia.
However, existing clinical studies have demonstrated a specific connection between the 2, which can provide some support for the results of this study.
This study aimed to reveal the correlation between GM and hyperlipidemia and to elucidate the potential connection between the 2. In this analysis, SNPs were screened using a series of quality control measures. The results demonstrated that orders, families, and genera had specific microbiota closely related to hyperlipidemia to a certain extent.
Nine human GM species were found to have a causal relationship with the risk of hyperlipidemia. Among these, Enterobacteriaceae, Adlercreutzia, LachnospiraceaeUCG001, Ruminococcus1, and Enterobacteriales had a risk effect on hyperlipidemia. Oxalobacteraceae, Clostridiumsensustricto1, RuminococcaceaeUCG010, and Rhodospirillales were protective factors against hyperlipidemia.
Sensitivity analyses revealed no horizontal pleomorphism, suggesting that MR analysis was not affected by confounding factors. The “leave one” analysis confirmed the robustness of this study.
This study identified that 5 intestinal microbiota (Enterobacteriaceae, Adlercreutzia, LachnospiraceaeUCG001, Ruminococcus1, and Enterobacteriales) may have a causal relationship with an increased risk of hyperlipidemia. In other words, the risk of hyperlipidemia increased with the abundance of these bacterial communities. Research on Enterobacteriaceae is scarce.
However, recent studies have revealed that these microbiota are closely linked to immune regulation. The growth of Enterobacteriaceae and Enterobacteriales is a biomarker of intestinal microbiota dysregulation in inflammatory bowel disease.[33-35] Consequently, it is hypothesized that by affecting the regulation of immune function, it may indirectly lead to the onset of hyperlipidemia. Adlercreutzia is a type of intestinal microbiota that regulates intestinal flora and metabolic activities, thereby alleviating inflammatory bowel disease and potentially interacting with antidepressant drugs.[36,37] Interestingly, research has found that an overabundance of Adlercreutzia may also be an essential factor in the onset of hyperlipidemia,[13,38,39] but the specific mechanism remains unclear.
LachnospiraceaeUCG001 and Ruminococcus1 promote the production of short-chain fatty acids and usually support metabolic health.[40,41] Studies have demonstrated that consuming a similar high-fat diet can increase LachnospiraceaeUCG001,[42] which is consistent with the findings of this study. It is further speculated that the presence of this microbiota is closely related to lipid metabolism and can directly or indirectly contribute to hyperlipidemia.
Furthermore, this study revealed that 4 GM (Oxalobacteraceae, Clostridiumsensustricto1, RuminococcaceaeUCG010, and Rhodospirillales) appeared to have protective effects against hyperlipidemia. In other words, the higher the abundance of this type of flora, the lower the risk of hyperlipidemia. Oxalobacteraceae, as a protective factor, has been found to regulate metabolic and neurological functions.[43,44] Consequently, these intestinal microbiota may play an important role in metabolic pathways.
Clostridium sensu stricto 1 can break down polysaccharides to produce short-chain fatty acids, promote the absorption of polysaccharides, and inhibit adipokines induced by intestinal hunger,[45,46] thereby increasing lipid metabolism[47] and lowering blood lipids. RuminococcaceaeUCG010 can produce short-chain fatty acids, such as butyrate and propionate, reduce the release of inflammatory factors, alleviate endothelial dysfunction, enhance the integrity of the intestinal barrier,[42] and further affect blood lipids. The abundance of Rhodospirillales, as one of the protective factors in this study, was found to have a negative correlation with cardiometabolism; however, the specific mechanism underlying this correlation is not yet clearly understood.[48] This study implies that increasing the abundance of Rhodospirillales is beneficial for reducing the incidence of hyperlipidemia.
As explained above, the composition and function of the intestinal microbiota may influence the occurrence of hyperlipidemia through multiple mechanisms; however, the majority of these mechanisms remain unclear. Clinical studies have found that observing 239 Japanese participants revealed a close correlation between the changes in the structure of the intestinal flora and alterations in hypertension and hyperlipidemia in the subjects.[49] Interestingly, studies have demonstrated that Lactobacillus acidophilus can restore the intestinal flora imbalance induced by a high-fat diet and maintain intestinal barrier integrity, thereby reducing metabolic endotoxemia and inhibiting the TLR4/NF-κB signaling pathway.[50] Notably, Lactobacillus plantarum HT121 can lower blood lipid levels[13,39,51] in rat models caused by a high-fat diet through changes in intestinal flora and bile acid metabolism. Further studies have pointed out that intestinal strains can produce short-chain fatty acids and secondary bile acids and affect the efficacy of statins by regulating the activity of proteins related to cholesterol metabolites.[52-54] Hyperlipidemia may reduce the source of nutrients in the intestine and change the reduction-oxidation reaction state of the intestine, thereby destroying the microenvironment on which the microbiota depend for survival.
Changes in this microenvironment may alter the composition of the intestinal microbiota and subsequently affect its ability to regulate blood lipids.[55-57] The intestinal microbiota regulates lipid metabolism, directly controls the expression of genes involved in fat storage in the host, and modulates the activity of polysaccharides, thereby influencing the body’s chronic low-level inflammatory state.[58,59] Notably, regulating the intestinal flora through methods, such as intestinal flora transplantation, can effectively improve abnormal blood lipid conditions. Additionally, supplementing prebiotics or probiotics is also an effective technique to regulate the intestinal flora and improve blood lipid levels.[60-62] There is no doubt that a dynamic interaction exists between the GM and hyperlipidemia,[63] which represents a new development for bacterium-based therapies in the future.
Limitations of the research: This study has several limitations, including the following: the data were acquired from the European region. It may not be universal due to factors such as dietary and lifestyle habits, genetic differences, and cultural background. Efforts should be made to include populations of all races for genetic research.
Currently, the research is unable to explore the entire lineage of the intestinal microbiota comprehensively. There may be omissions in the middle, particularly regarding the microbiota, which can increase the risk. Since this study is a data summary, it is only able to prove the causal relationship between GM and hyperlipidemia.
It is unable to statistically analyze the information to conduct a more specific stratified analysis of the sample. Additional multicenter, large-sample, high-quality observational and clinical trials are still needed.
Future studies should be designed to explore how the interaction between host genetic factors and GM jointly affects the risk of hyperlipidemia, as well as the regulatory effects of environmental and lifestyle factors (such as diet and exercise) on the link between intestinal microbiota and hyperlipidemia. A large-scale MR study was conducted among different races to verify the universality of the results of this study, and a long-term longitudinal study was designed to evaluate the time-dependent effect of GM on hyperlipidemia development.
5. Conclusion
This study used the 2-sample MR method to confirm the potential causal relationship between hyperlipidemia and intestinal microbiota, providing a theoretical basis for treating and preventing hyperlipidemia through targeted regulation of intestinal microbiota. It also offered new perspectives and targets for the pathogenesis and treatment of hyperlipidemia. However, this study did not clearly define the specific role patterns of these intestinal microbiota in hyperlipidemia pathogenesis.
Future studies should focus on thoroughly elucidating the causal mechanism and exploring customized diagnosis and treatment strategies based on microbiota characteristics in conjunction with the host’s genetic background, dietary habits, lifestyle, and other factors. This strategy will enable the development of safe and effective targeted intervention methods and the verification of their long-term benefits through rigorous clinical research. Large-scale MR studies were conducted among different racial groups to verify the universality of the results of this study and pave the way for the precise prevention and control of hyperlipidemia by the intestinal microbiota.
Acknowledgments
All the statistics can be found in Genome-wide association study (GWAS) meta-analysis (http://www.mibiogen.org/) and IEU OpenGWAS project (https://gwas.mrcieu.ac.uk/).
Author contributions
Formal analysis: Changpeng Du, Yingru Chen.
Visualization: Changpeng Du, Yingru Chen, Xijin Wei, Zhen Wang.
Writing – original draft: Changpeng Du, Yingru Chen, Yan Cheng.
Writing – review & editing: Changpeng Du, Yingru Chen, Zhen Wang.