Your privacy, your choice

We use essential cookies to make sure the site can function. We also use optional cookies for advertising, personalisation of content, usage analysis, and social media.

By accepting optional cookies, you consent to the processing of your personal data - including transfers to third parties. Some third parties are outside of the European Economic Area, with varying standards of data protection.

See our privacy policy for more information on the use of your personal data.

for further information and to change your choices.
Skip to main content
Download PDF

Unraveling the causal pathway between phosphatidylinositol, metabolites, and metabolic syndrome: a Mendelian randomization study

Diabetology & Metabolic Syndrome volume 17, Article number: 162 (2025) Cite this article

Abstract

Introduction

Observational studies have increasingly acknowledged the influence of Phosphatidylinositol (PI) on metabolic syndrome (MetS). Nevertheless, the causal association between PI and MetS remains unclear due to the presence of confounding factors and the potential for reverse causation in observational settings. This study seeks to clarify the causal link between PI and MetS while investigating the role of mediating metabolites.

Methods

A two-sample Mendelian randomization (MR) analysis was performed to examine the association between PI and MetS, utilizing aggregated data from genome-wide association studies (GWAS). Additionally, a two-step MR approach was applied to quantify the mediation effect of metabolites on the PI-MetS relationship. The inverse variance weighted (IVW) method served as the primary analytical approach, complemented by various sensitivity analyses employing alternative techniques.

Results

A significant positive association was found between genetically predicted PI and a 17% increased risk of MetS. Genetically predicted metabolites, including 4-cholesten-3-one (IVW: OR 1.264, 95% CI 1.076–1.483, p = 0.004), N-acetylalliin (IVW: OR 1.189, 95% CI 1.008–1.402, p = 0.040), and the Adenosine 5’-diphosphate to 5-oxoproline ratio (IVW: OR 1.191, 95% CI 1.045–1.357, p = 0.009), were each significantly associated with an increased risks of MetS, accounting for 14.50, 11.41%, 11.87% and % of the total effect, respectively. Notably, the Retinol to oleoyl-linoleoyl-glycerol ratio (IVW: OR 0.643, 95% CI 0.466–0.887, p = 0.007) mediated 62.6% of the effect, highlighting its pivotal role in the causal pathway linking PI to MetS. Moreover, 1-palmitoyl-2-dihomo-linolenoyl-GPC (IVW: OR 0.865, 95% CI 0.752–0.995, p = 0.042) and the Creatine to carnitine ratio (IVW: OR 0.853, 95% CI 0.740–0.983, p = 0.028) were associated with a reduced risk of MetS, demonstrating inhibitory effects within their respective pathways that accounted to 35.03% and 8.45% reductions in risk, respectively.

Conclusions

Our MR analysis demonstrated a positive association between PI and an increased risk of MetS. Furthermore, the metabolite-mediated PI significantly influenced MetS risk. These findings may offer valuable insights into the pathogenesis of MetS and inform future clinical research.

Introduction

Metabolic syndrome (MetS) is characterized by a constellation of metabolic abnormalities, including dyslipidemia, hypertension, obesity, and insulin resistance, all of which synergistically increase the risk of cardiovascular disease (CVD) [1, 2]. In recent years, the prevalence of MetS has persistently increased, posing a significant public health challenge on a global scale. Epidemiological studies reveal that MetS affects approximately 6–39% of adolescents globally, with prevalence rates in adults ranging from 20–25% [3, 4]. These findings underscore the profound impact of MetS as a widespread public health issue, placing a substantial burden on healthcare systems worldwide. Therefore, a comprehensive investigation into the pathological mechanisms underlying MetS, coupled with the development of potential therapeutic strategies, is of significant scientific and clinical importance.

Phosphatidylinositol (PI) is an essential phospholipid in the human body that produces seven distinct phosphorylated inositol lipids [5]. It functions as a fundamental component of the cell membrane, significantly contributing to cell signaling and membrane transport [6, 7]. PI is intricately linked with insulin resistance, obesity, inflammation, and other critical aspects of MetS, affecting glucose-lipid metabolism through various pathways that facilitate the progression of MetS [8, 9]. Conversely, emerging studies propose that PI might also regulate lipid metabolism and energy homeostasis, potentially exerting an anti-obesity effect by modulating the expression of specific genes involved in hepatic lipid metabolism through diverse metabolic pathways [8]. Derivatives of PI, such as PI3, may serve as insulin sensitizers, offering potential therapeutic benefits in ameliorating metabolic disorders [10,11,12]. The complex mechanisms through which PI influences MetS underscore the intricate nature of their relationship, with the causal link between them remaining to be fully elucidated.

Previous research has established that metabolites function as pivotal signaling molecules in the pathological processes underlying MetS, impacting mechanisms such as energy homeostasis, glucose metabolism, lipid metabolism, and inflammatory regulation [12, 13]. Phosphatidylinositol 3-phosphate (PI3P) and Phosphatidylinositol 5-phosphate (PI5P), which are principal derivatives of PI, are acknowledged for their regulatory interactions with a broad spectrum of metabolites [14,15,16,17]. Consequently, it is apparent that these metabolites are not merely by-products of metabolic processes but also play significant roles in regulating both the onset and progression of MetS. Based on these observations, we hypothesize that metabolites might mediate the effects of PI on MetS.

Mendelian Randomization (MR) is a statistical approach that utilizes genetic variants as instrumental variables (IVs) to mimic the randomization process of controlled clinical trials. This approach inherently mitigates confounding factors such as environmental and behavioral variables, and effectively overcomes the issue of reverse causality [18, 19].Consequently, MR has become a widely accepted strategy for establishing causal inferences in complex disease research. In this study, we applied MR to examine the causal relationships between PI and MetS, and further assessed the mediating effects of candidate metabolic intermediates.

Methods

Study design

To elucidate the causal relationship between PI and MetS, this study employed large-scale, publicly available genome-wide association study (GWAS) pooled statistics. This MR study adhered to the Strengthening the Reporting of Observational Studies in Epidemiology using Mendelian Randomization (STROBE-MR) guidelines [20].

The study was conducted in three stages. In the initial stage, a univariate Mendelian randomization study (UVMR) was performed to examine the causal relationship between PI and MetS, identifying a positive effect of PI on MetS. To further verify the causal direction and exclude the potential for reverse causation, we conducted a reverse MR analysis to evaluate the influence of MetS on PI. The analysis yielded no evidence of a significant association between MetS and PI. This analytical approach reinforces the robustness of the causal inference and further substantiates the validity of our findings. In the second phase, the causal relationship between PI and metabolites was assessed, resulting in the identification of 81 metabolites as potential mediators. Finally, the impact of metabolites on MetS was examined, leading to the screening of six metabolites and the application of a two-step MR approach to determine the mediator ratio of the compliant metabolites. The MR analysis was conducted under three key assumptions: a strong association between IVs and exposure factors (correlation assumption), the independence of IVs from confounders of the exposure-outcome relationship (independence assumption), and the effect of IVs on the outcome occurring solely through exposure (exclusivity assumption) [21].

Data source

PI summary statistics were obtained from FinnGen, which includes 7174 (GeneRISK’s 7174 Finns) sample populations of European ancestry [22].The FinnGen study is a large-scale genomics initiative that has analyzed over 500,000 Finnish biobank samples and correlated genetic variation with health data to understand disease mechanisms and predispositions. The project is a collaboration between research organisations and biobanks within Finland and international industry partners.

The GWAS summary statistics for MetS were obtained from The GWAS catalog (GCST90086073). The data was collected from 56,637 individuals of European descent using the comprehensive GWAS strategy GUIDANCE, leading to the identification of 94 genome-wide relevant loci, 26 of which were previously unreported [23]. Metabolite data were sourced from the Canadian Longitudinal Study of Aging (CLSA), which involved 8,299 participants [24]. The CLSA is a large-scale study that encompasses data from 50,000 participants, contributing to a more profound understanding of the disease. Both the GWAS and the CLSA provided the metabolite data used in this study.

Although the PI, MetS, and metabolite data were derived from individuals of European ancestry, differences in sample size, study design, and data acquisition protocols may contribute to population heterogeneity. To minimize the influence of population stratification, only GWAS summary statistics derived from European-ancestry samples were included in the variable selection process. Sensitivity analyses addressing potential bias from population stratification were performed using inverse variance weighting (IVW) and MR-Egger methods, both of which indicated no substantial heterogeneity. Additional information regarding the datasets is provided in Supplementary Table S1.

Genetic instrumental variable selection

Given the limited number of genome-wide significant variants linked to PI and specific metabolites, we selected SNPs associated with PI, metabolites, and MetS from GWAS datasets using a significance threshold of P < 1 × 10⁻⁵. To ensure the independence of selected instrumental variables, we applied a stringent linkage disequilibrium (LD) clumping strategy using a window size of 10,000 base pairs and an r² threshold of < 0.001. This approach effectively minimizes covariance-induced bias in MR estimates and addresses instrument non-independence caused by LD, thereby reducing the risk of false-positive results. We then calculated the F-statistic for each SNP and excluded those classified as weak IVs (F < 10). In addition, allelic harmonization was performed to align effect alleles and prevent strand mismatches, thereby improving the accuracy of causal inference. To account for potential pleiotropic bias, we conducted sensitivity analyses using methods such as MR-Egger and MR-PRESSO to detect horizontal pleiotropy and confirm the robustness of our findings. Detailed characterization of SNPs associated with PI, Metabolites, and MetS, respectively, can be found in the online Supplementary Material, Tables S2-S5.

Statistical analysis

UVMR analysis

In the UVMR analysis, the IVW method was primarily used to estimate causal effects. This approach provides the highest statistical power under the assumption that all instrumental variables are valid and not affected by horizontal pleiotropy. Recognizing the potential for pleiotropic bias in empirical settings, we additionally employed MR-Egger regression, the weighted median method, as well as the simple and weighted mode-based estimators to validate the robustness of causal inference. The weighted median method yields consistent causal estimates even when up to 50% of the instruments are invalid, whereas MR-Egger regression enables detection and adjustment of directional pleiotropy [25, 26]. MR-Egger intercept tests were further used to assess global pleiotropy, and heterogeneity was quantified using Cochran’s Q statistic. Additionally, MR-PRESSO was utilized to detect potential outlier IVs and correct for any bias they may introduce. The reliability of IVW-derived estimates is reinforced when results from multiple methods demonstrate directional concordance and sensitivity analyses reveal no evidence of substantial pleiotropy or heterogeneity.

Mediator MR analysis

In this study, a two-step MR analysis was conducted to investigate potential intermediate factors that may mediate the causal relationship between PI and MetS [27]. Initially, the study evaluated the causal impact of PI on selected mediators (β1), followed by an assessment of the mediators’ causal effect on MetS (β2). Subsequently, utilizing the ‘product of coefficients method,’ the study calculated the mediating effect of PI on MetS through the mediator (β1*β2) and determined the mediation ratio by dividing the mediating effect by the total effect (the causal effect of PI on MetS) (Fig. 1).

Fig. 1
figure 1

Study protocol

Full size image

Sensitivity analysis

In this study, we assessed the robustness of the IVW results using MR-Egger, weighted median, simple mode, and weighted mode. Cochran’s Q statistic and funnel plots were used to test for heterogeneity [21, 28].MR-Egger assesses the multiplicity of genetic variants based on the magnitude of the intercept term [29]. We used the MR polytomous residuals and outliers (MR-PRESSO) method to detect possible outliers and calculated causal estimates after removing identified outliers [30]. In addition, forest plots and leave-one-out plots were drawn for sensitivity analysis. All MR analyses were performed using the R packages “Two Sample MR”, “MR. raps” in R software (version 4.2.3; https://www.r-project.org), “MR-PRESSO”.

Results

Causal relationship between PI and MetS

A bidirectional MR analysis was conducted to investigate the relationship between PI and MetS. When PI was considered as the exposure, nine single nucleotide polymorphisms (SNPs) were selected as IVs for the MR analysis (see Supplementary Table S2). The primary IVW analysis indicated a positive association between PI and the risk of MetS, revealing a 17% increase in MetS risk for each 1 standard deviation increase in PI (OR: 1.173; 95% CI: 1.039–1.324). These findings were further confirmed by other sensitivity analysis methods. No heterogeneity among the IVs was detected, as indicated by Cochran’s Q statistic, and instrument validity tests confirmed strong instrument strength (F > 10) with no evidence of horizontal pleiotropy. Funnel plot analysis did not reveal any asymmetry. Results from the leave-one-out analysis demonstrated that causality estimates remained stable, with no significant changes observed after the exclusion of any single SNP. When MetS was treated as the exposure, thirty-three SNPs associated with MetS were identified (see Supplementary Table S3). However, substantial evidence indicated no association between MetS and PI, with the IVW estimate being 0.972 (95% CI: 0.900–1.050, P = 0.475) (Fig. 2).

Fig. 2
figure 2

Causal effect of PI on MetS

Full size image

Causal relationship between PI and metabolites

A total of 81 out of 1400 metabolites met all initial screening criteria (Fig. 3). These IVW estimates have been confirmed by at least one sensitivity analysis. Possible heterogeneity between some of the IVs in this study. the MR-Egger pleiotropy test showed horizontal pleiotropy between PI and metabolites such as 1-palmitoyl-GPI, while the MR-PRESSO method was used to exclude isolated SNPs. After excluding these outlier SNPs, IVW estimates for the majority of metabolites remained largely unchanged. Minor adjustments were observed for a few metabolites (e.g., ROL/GLY), although the direction of the effects was preserved. These findings underscore the robustness of our analytical approach. Furthermore, although slight asymmetry was observed in a subset of funnel plots, most remained directionally consistent with IVW estimates. Causal effect estimates also showed minimal variation during the stepwise removal of individual SNPs (Supplementary Tables S4–S7).

Fig. 3
figure 3

The selection process of Metabolites as causal mediators of PI and MetS

Full size image

Causality of metabolites on MetS

Among the 81 metabolites assessed, six metabolites fulfilled all screening criteria. Notably, 4-cholesten-3-one (4C3O), N-acetylalliin (NAA), and the ratio of adenosine 5’-diphosphate to 5-oxoproline (ADP/OP) demonstrated a positive correlation with the risk of MetS, leading to a 18–26% increase in risk per 1 SD increase. Conversely, 1-palmitoyl-2-dihomo-linolenoyl-GPC (PDL-GPC), the creatine to carnitine ratio (CCR), and the retinol to oleoyl-linoleoyl-glycerol ratio (ROL/GLY) were associated with a 13–36% reduction in MetS risk per 1 standard deviation (SD) increase. In sensitivity analyses of metabolite causality with MetS, the MR-Egger intercept method revealed no significant directed pleiotropy in any of the analyses. Analysis using Cochran’s Q statistic revealed potential heterogeneity among the partial IVs. Additionally, deviations from symmetry were noted in several funnel plots. Following the exclusion of potential pleiotropic SNPs, the MR-PRESSO method produced results consistent with the IVW analysis, with the exception of ROL/GLY (supplementary material online, document 1).

Mediating role of metabolites between PI and MetS

This study identified that ROL/GLY (mediating effect percentage: 62.59%), 4C3O (14.50%), ADP/OP (11.87%), and NAA (11.41%) mediated the effect of PI on MetS. Furthermore, the study demonstrated that PDL-GPC and CCR exerted inhibitory effects on the promotion of MetS by PI. Specifically, PDL-GPC exhibited a 35.03% inhibitory effect, whereas CCR demonstrated an 8.45% inhibitory effect (Figs. 4 and 5).

Fig. 4
figure 4

Causal effect of PI on metabolites

Full size image
Fig. 5
figure 5

Causal effect of metabolites on MetS

Full size image

Discussion

This MR study reveals a positive causal effect of PI on MetS and elucidates the mediating pathways involved. We observed that each standard deviation (SD) increase in genetically determined PI corresponded to a 17% increase in MetS risk. Concurrently, we identified that 4C3O, NAA, ADP/OP, and ROL/GLY facilitated PI-promoted MetS development, with ROL/GLY exerting the most substantial effect at 62.59%. Notably, PDL-GPC and the CCR played an inhibitory role in this pathway.

PI is synthesized from inositol and phosphatidic acid in the endoplasmic reticulum, serving as a minor phospholipid closely linked to insulin levels. Its phosphorylation acts as a crucial cell signaling mechanism involved in various biological processes like cell growth, differentiation, migration, and apoptosis. PI can be phosphorylated at different positions on the inositol ring, generating diverse phosphorylation products such as PI3P, PI4P, PI5P, PI (3,4)P2, PI(4,5)P2, and PI(3,4,5)P3, each with distinct biological functions [5, 30, 31]. For instance, PI3P is pivotal in activating the PI3K/Akt signaling pathway, which influences the homeostatic and metabolic functions of adipose tissue [32]. In cases of obesity and MetS, reduced cellular response to insulin can inhibit the PI3K/AKT pathway, impacting glucose and lipid metabolism. Additionally, MetS often involves low-grade systemic inflammation, where the PI signaling pathway contributes to the inflammatory response, potentially leading to increased production of inflammatory factors that block the IRS-PI3K/AKT pathway, worsening metabolic disturbances and contributing to the prevalence of MetS [33, 34]. PI4P as one of its derivatives, is important for maintaining many cellular functions and has been closely associated with diseases such as obesity and type 2 diabetes [35, 36]. PI(4,5)P2 is an interesting and important phospholipid in the human body, which is mainly enriched in the plasma membrane and affects the function of organelles by regulating the localization and activity of proteins on the organelles and has an important role in the development of diabetes mellitus [37, 38]. Verma et al. [39]. highlighted the pivotal role of the PI3K/AKT pathway in MetS progression, showing its association with lipid levels, atherosclerosis, and insulin levels. Phosphorylated PI derivatives serve as intracellular signaling molecules, interacting with proteins through specific structural domains like PH, FYVE, and PX domains, thereby transmitting signals within the cell. Ratke et al. [40]. investigated the signaling mechanism behind leptin-induced tumor cell migration and identified PI3K activation by leptin as a key factor in this process, which has implications for the management of obesity-related colorectal cancer. Using MR methods, the relationship between PI and MetS was explored, revealing that elevated PI levels contribute to MetS development, underscoring the importance of long-term PI effects on MetS and offering potential clinical strategies for its prevention and treatment.

Previous research has established a strong relationship between metabolites, MetS, and PI, emphasizing their crucial role in the development of metabolic disorders [13]. This study seeks to identify potential metabolite mediators that influence the causal effects of PI on the progression of MetS. The analysis identified that the ratio of retinol to glyceryl (ROL/GLY) predominantly mediates the pathway through which elevated PI levels increase the risk of MetS, accounting for 62.59% of the effect. Retinol has been demonstrated to affect the phosphoinositide 3-kinase (PI3K)/AKT signaling pathway, which plays a role in vascular endothelial function and neuroprotection [41, 42]. Additionally, elevated levels of retinol have been associated with insulin resistance, increased triglyceride (TG) levels, and an elevated risk of type 2 diabetes [43]. Further, retinol-binding protein 4 (RBP4) has been implicated in the development of hypertriglyceridemia and MetS in multiple studies [44]. Reay et al. [45]. reported that retinol may causally influence inflammation, obesity, ocular phenotypes, gut microbiota composition, and MRI-derived brain structures. Conversely, circulating retinol levels may themselves be causally influenced by lipid profiles and serum creatinine concentrations. Shin et al. [46]. demonstrated that cytosolic retinol homeostasis dysregulation is closely linked to RBP4 overproduction and hepatic inflammation in diabetic livers, ultimately impairing hepatic metabolic function. Moreover, Tian et al. [47]. found that exogenous retinyl acetate remodels the cellular lipidome by elevating phosphatidylcholine and TG levels while reducing PI content. Collectively, these findings support the potential utility of the ROL/GLY ratio as an early biomarker for metabolic dysregulation.

Another notable mediator identified was 4C3O, contributing 14.50% to the causal pathway. This cholesterol oxidation product, predominantly metabolized in the liver, interacts with motor proteins in conjunction with PI, demonstrating anti-obesity and lipid-lowering effects [48]. Elia et al. [49]. demonstrated that 4C3O exerts antitumor effects by modulating lipid metabolism and suppressing adipogenesis in breast cancer cells. Liu et al. [50]. conducted a meta-analysis and systematic review of ten studies involving 1,034 patients with irritable bowel syndrome and 232 healthy controls. The analysis revealed significantly elevated serum 4C3O levels in irritable bowel syndrome patients compared to healthy individuals. Delle Bovi RJ et al. [51]. demonstrated that elevated levels of 4C3O could further exacerbate metabolic abnormalities by influencing cell membrane structure or insulin receptor function. Furthermore, a cohort study by Pang et al. [52]. found that metabolic markers, including 4C3O, were associated with the risk of non-alcoholic fatty liver disease in relatively lean Chinese adults, suggesting a potential role for these biomarkers in mediating the link between metabolic dysregulation and non-alcoholic fatty liver disease.

Notably, our study also demonstrated that PDL-GPC and CCR significantly inhibit the development of MetS exacerbated by elevated PI levels. Creatine plays a role in antioxidant and anti-inflammatory responses and supports metabolic processes. Conversely, fluoride has been demonstrated to decrease levels of creatine and other metabolites by activating signaling pathways such as FoxO and PI3K/Akt. In conclusion, our research highlights the critical role of metabolites in mediating the promotion of MetS by elevated PI levels. These findings suggest that modifying dietary habits and lifestyle to regulate metabolite levels of retinol, creatine, and other compounds may help mitigate the impact of elevated PI levels on MetS.

Additionally, we identified that NAA and the ADP/OP mediates the causal relationship between PI and MetS, thereby contributing to disease progression. NAA, a sulfur-containing compound, is potentially associated with antioxidant properties. Its elevation may be implicated in the development of insulin resistance through the regulation of oxidative stress and inflammatory responses. Insulin resistance is commonly associated with chronic inflammation, and elevated oxidative stress may exacerbate this condition, thus contributing to the development of MetS [53]. Increased ADP levels are linked to a reduced cellular energy supply, whereas the metabolism of 5-oxoproline may be associated with antioxidant functions within the glutathione cycle. Additionally, NAA and ADP/OP were found to mediate the relationship between PI and the progression of MetS [54, 55]. The PI3K-mediated Kkt1 pathway plays a pivotal role in the release of platelet al.pha granules induced by ADP, thereby affecting platelet function during the inflammatory response. Consequently, alterations in this ratio may indicate dysregulation of metabolic and antioxidant defense systems, further influencing the effects of PI on MetS [56, 57]. Research has indicated that the Kkt1 pathway, mediated by PI3K, is crucial for ADP-induced platelet α-granule release and plays a significant role in orchestrating platelet-mediated inflammatory responses [58].

Interestingly, we also observed that the ratio of 1-palmitoyl-2-dihomo-linolenoyl-GPC to creatine and carnitine plays a significant role in inhibiting the development of MetS accelerated by PI. This finding suggests that specific molecules associated with lipid and energy metabolism may play a role in the metabolic protective mechanism. The creatine to carnitine ratio is involved in antioxidant and anti-inflammatory responses that support metabolic processes. Additionally, fluoride decreases the levels of creatine and other metabolites by activating signaling pathways such as FoxO and PI3K/Akt [59]. By enhancing cellular energy metabolism, creatine, as a component of the energy buffer system, improves ATP regeneration efficiency, while carnitine is essential for fatty acid oxidation. The synergistic effects of these compounds may inhibit PI-induced metabolic disorders by improving energy supply and fatty acid metabolism [60].

In conclusion, our findings support a positive causal relationship between PI and MetS, offering new insights into metabolic regulatory networks mediated by diverse biomarkers. While elucidating the underlying mechanisms of PI action in metabolic diseases, these findings highlight several potential targets for intervention, particularly in retinol metabolism, cholesterol metabolism, and energy metabolism pathways. However, we acknowledge that the robustness of our results may be influenced by the inherent limitations of the MR methodology, including the limited generalizability due to reliance on European population samples and the inability of MR to capture potential nonlinear relationships. Consequently, these results should be interpreted with caution and necessitate validation in both experimental and population-based settings. Future research should continue to investigate the mechanisms of these metabolic markers in metabolic disorders, aiming to provide a more robust scientific foundation for prevention and treatment strategies for MetS.

A major strength of this study lies in the implementation of a rigorously designed MR approach, leveraging genetic variants as instrumental variables to investigate the potential causal relationship between PI and MetS, and to further examine possible metabolite-mediated pathways. We utilized large-scale, publicly available GWAS summary statistics for genetic instruments related to exposures, outcomes, and mediators, with rigorous allelic harmonization to ensure strand and effect alignment. Comprehensive sensitivity analyses including MR-Egger regression, weighted median and mode estimators, and MR-PRESSO were conducted to assess potential horizontal pleiotropy and evaluate the robustness of the causal inferences. Nonetheless, this study has several notable limitations. First, our focus on individuals of European ancestry was aimed at ensuring genetic consistency, which may restrict the generalizability of our findings to other racial and ethnic groups. Second, given the two-sample MR design, we assumed a linear relationship between PI and MetS; thus, individual-level data will be essential for future research to investigate potential nonlinear causality. Third, although we employed methods to address abnormal variability, the potential impact of horizontal pleiotropy and the heterogeneity of IVs on our results cannot be fully excluded. Fourth, due to the use of pooled statistics in this study, we are unable to determine causal relationships within subgroups, such as between females and males. Fifth, the limited data on PIs and metabolites within the GWAS datasets underscores the need for more comprehensive data in future studies to facilitate further validation. Finally, this study employed a two-step MR approach to perform mediation analysis, estimating the proportion mediated via the “product of coefficients” method. However, this approach generally assumes theoretically that there is no interaction between exposure and the mediating variable, which may not always hold true in real biological mechanisms. Although we initially assessed potential interaction effects in our sensitivity analyses, the current analyses failed to systematically incorporate modeling of the interaction terms, which may have led to underestimation or bias of the mediating effects.

Conclusion

The study demonstrated a positive causal relationship between PI and MetS, with key mediators identified as the ROL/GLY, 4C3O, NAA, and ADP/OP. Conversely, PDL-GPC and the CCR were found to play an inhibitory role in this pathway. These findings offer new insights into the pathogenesis of MetS, highlighting the importance of dietary and weight management in lipid regulation.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

References

  1. Hsu CN, Hou CY, Hsu WH, Tain YL. Early-Life Origins of Metabolic Syndrome: Mechanisms and Preventive Aspects. International journal of molecular sciences 2021, 22(21).

  2. Fahed G, Aoun L, Bou Zerdan M, Allam S, Bou Zerdan M, Bouferraa Y, Assi HI. Metabolic syndrome: updates on pathophysiology and management in 2021. Int J Mol Sci 2022, 23(2).

  3. Weihe P, Weihrauch-Blüher S. Metabolic syndrome in children and adolescents: diagnostic criteria, therapeutic options and perspectives. Curr Obes Rep. 2019;8(4):472–9.

    Article  PubMed  Google Scholar 

  4. Alkhulaifi F, Darkoh C. Meal timing, meal frequency and metabolic syndrome. Nutrients 2022, 14(9).

  5. Ashlin TG, Blunsom NJ, Cockcroft S. Courier service for phosphatidylinositol: pitps deliver on demand. Biochim Et Biophys Acta Mol Cell Biology Lipids. 2021;1866(9):158985.

    Article  CAS  Google Scholar 

  6. Nakada-Tsukui K, Watanabe N, Maehama T, Nozaki T. Phosphatidylinositol kinases and phosphatases in Entamoeba histolytica. Front Cell Infect Microbiol. 2019;9:150.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Dickson EJ, Hille B. Understanding phosphoinositides: rare, dynamic, and essential membrane phospholipids. Biochem J. 2019;476(1):1–23.

    Article  CAS  PubMed  Google Scholar 

  8. Shimizu K, Ida T, Tsutsui H, Asai T, Otsubo K, Oku N. Anti-obesity effect of phosphatidylinositol on diet-induced obesity in mice. J Agric Food Chem. 2010;58(21):11218–25.

    Article  CAS  PubMed  Google Scholar 

  9. Le Stunff C, Dechartres A, Miraglia Del Giudice E, Froguel P, Bougnères P. A single-nucleotide polymorphism in the p110beta gene promoter is associated with partial protection from insulin resistance in severely obese adolescents. J Clin Endocrinol Metab. 2008;93(1):212–5.

    Article  PubMed  Google Scholar 

  10. Croze ML, Soulage CO. Potential role and therapeutic interests of myo-inositol in metabolic diseases. Biochimie. 2013;95(10):1811–27.

    Article  CAS  PubMed  Google Scholar 

  11. Lee N, Kim I, Park S, Han D, Ha S, Kwon M, Kim J, Byun SH, Oh W, Jeon HB, et al. Creatine inhibits adipogenesis by downregulating insulin-induced activation of the phosphatidylinositol 3-kinase signaling pathway. Stem Cells Dev. 2015;24(8):983–94.

    Article  CAS  PubMed  Google Scholar 

  12. Hoshino F, Murakami C, Sakai H, Satoh M, Sakane F. Creatine kinase muscle type specifically interacts with saturated fatty acid- and/or monounsaturated fatty acid-containing phosphatidic acids. Biochem Biophys Res Commun. 2019;513(4):1035–40.

    Article  CAS  PubMed  Google Scholar 

  13. Wu Q, Li J, Sun X, He D, Cheng Z, Li J, Zhang X, Xie Y, Zhu Y, Lai M. Multi-stage metabolomics and genetic analyses identified metabolite biomarkers of metabolic syndrome and their genetic determinants. EBioMedicine. 2021;74:103707.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Li CZ, Wu LM, Zhu CX, Du HY, Chen GX, Yang F. The impacts of dietary sphingomyelin supplementation on metabolic parameters of healthy adults: a systematic review and meta-analysis of randomized controlled trials. Front Nutr. 2024;11:1363077.

    Article  PubMed  PubMed Central  Google Scholar 

  15. Vaillant F, Corrales-Agudelo V, Moreno-Castellanos N, Ángel-Martín A, Henao-Rojas JC, Muñoz-Durango K, Poucheret P. Plasma Metabolome Profiling by High-Performance Chemical Isotope-Labelling LC-MS after Acute and Medium-Term Intervention with Golden Berry Fruit (Physalis peruviana L.), Confirming Its Impact on Insulin-Associated Signaling Pathways. Nutrients 2021, 13(9).

  16. Savova MS, Mihaylova LV, Tews D, Wabitsch M, Georgiev MI. Targeting PI3K/AKT signaling pathway in obesity. Biomed pharmacotherapy = Biomedecine Pharmacotherapie. 2023;159:114244.

    Article  CAS  PubMed  Google Scholar 

  17. Ding E, Deng F, Fang J, Li T, Hou M, Liu J, Miao K, Yan W, Fang K, Shi W, et al. Association between organophosphate ester exposure and insulin resistance with glycometabolic disorders among older Chinese adults 60–69 years of age: evidence from the China BAPE study. Environ Health Perspect. 2023;131(4):47009.

    Article  CAS  PubMed  Google Scholar 

  18. Park EY, Wilder ET, Chipuk JE, Lane MA. Retinol decreases phosphatidylinositol 3-kinase activity in colon cancer cells. Mol Carcinog. 2008;47(4):264–74.

    Article  CAS  PubMed  Google Scholar 

  19. Liu C, Zhou XR, Ye MY, Xu XQ, Zhang YW, Liu H, Huang XZ. RBP4 is associated with insulin resistance in hyperuricemia-Induced rats and patients with hyperuricemia. Front Endocrinol. 2021;12:653819.

    Article  Google Scholar 

  20. Skrivankova VW, Richmond RC, Woolf BAR, Yarmolinsky J, Davies NM, Swanson SA, VanderWeele TJ, Higgins JPT, Timpson NJ, Dimou N, et al. Strengthening the reporting of observational studies in epidemiology using Mendelian randomization: the STROBE-MR statement. JAMA. 2021;326(16):1614–21.

    Article  PubMed  Google Scholar 

  21. Skrivankova VW, Richmond RC, Woolf BAR, Davies NM, Swanson SA, VanderWeele TJ, Timpson NJ, Higgins JPT, Dimou N, Langenberg C, et al. Strengthening the reporting of observational studies in epidemiology using Mendelian randomisation (STROBE-MR): explanation and elaboration. BMJ (Clinical Res ed). 2021;375:n2233.

    Google Scholar 

  22. Ottensmann L, Tabassum R, Ruotsalainen SE, Gerl MJ, Klose C, Widén E, Simons K, Ripatti S, Pirinen M. Genome-wide association analysis of plasma lipidome identifies 495 genetic associations. Nat Commun. 2023;14(1):6934.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Guindo-Martínez M, Amela R, Bonàs-Guarch S, Puiggròs M, Salvoro C, Miguel-Escalada I, Carey CE, Cole JB, Rüeger S, Atkinson E, et al. The impact of non-additive genetic associations on age-related complex diseases. Nat Commun. 2021;12(1):2436.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Chen Y, Lu T, Pettersson-Kymmer U, Stewart ID, Butler-Laporte G, Nakanishi T, Cerani A, Liang KYH, Yoshiji S, Willett JDS, et al. Genomic atlas of the plasma metabolome prioritizes metabolites implicated in human diseases. Nat Genet. 2023;55(1):44–53.

    Article  PubMed  PubMed Central  Google Scholar 

  25. Kong L, Wang Y, Ye C, Dou C, Liu D, Xu M, Zheng J, Zheng R, Xu Y, Li M, et al. Opposite causal effects of birthweight on myocardial infarction and atrial fibrillation and the distinct mediating pathways: a Mendelian randomization study. Cardiovasc Diabetol. 2023;22(1):338.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  26. Hartwig FP, Davey Smith G, Bowden J. Robust inference in summary data Mendelian randomization via the zero modal Pleiotropy assumption. Int J Epidemiol 2017, 46(6):1985–98.

  27. Carter AR, Sanderson E, Hammerton G, Richmond RC, Davey Smith G, Heron J, Taylor AE, Davies NM, Howe LD. Mendelian randomisation for mediation analysis: current methods and challenges for implementation. Eur J Epidemiol. 2021;36(5):465–78.

    Article  PubMed  PubMed Central  Google Scholar 

  28. Bowden J, Holmes MV. Meta-analysis and Mendelian randomization: A review. Res Synthesis Methods. 2019;10(4):486–96.

    Article  Google Scholar 

  29. Burgess S, Bowden J, Fall T, Ingelsson E, Thompson SG. Sensitivity analyses for robust causal inference from Mendelian randomization analyses with multiple genetic variants. Epidemiol (Cambridge Mass). 2017;28(1):30–42.

    Article  Google Scholar 

  30. Verbanck M, Chen CY, Neale B, Do R. Detection of widespread horizontal Pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet. 2018;50(5):693–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Blunsom NJ, Cockcroft S. Phosphatidylinositol synthesis at the Endoplasmic reticulum. Biochim Et Biophys Acta Mol Cell Biology Lipids. 2020;1865(1):158471.

    Article  CAS  Google Scholar 

  32. Huang X, Liu G, Guo J, Su Z. The PI3K/AKT pathway in obesity and type 2 diabetes. Int J Biol Sci. 2018;14(11):1483–96.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Acosta-Martinez M, Cabail MZ. The PI3K/Akt pathway in Meta-Inflammation. Int J Mol Sci 2022, 23(23).

  34. Tsay A, Wang JC. The role of PIK3R1 in metabolic function and insulin sensitivity. Int J Mol Sci 2023, 24(16).

  35. Marković V, Jaillais Y. Phosphatidylinositol 4-phosphate: a key determinant of plasma membrane identity and function in plants. New Phytol. 2022;235(3):867–74.

    Article  PubMed  Google Scholar 

  36. Li YP, Mikrani R, Hu YF, Faran Ashraf Baig MM, Abbas M, Akhtar F, Xu M. Research progress of phosphatidylinositol 4-kinase and its inhibitors in inflammatory diseases. Eur J Pharmacol. 2021;907:174300.

    Article  CAS  PubMed  Google Scholar 

  37. Wen Y, Vogt VM, Feigenson GW. PI(4,5)P(2) clustering and its impact on biological functions. Annu Rev Biochem. 2021;90:681–707.

    Article  CAS  PubMed  Google Scholar 

  38. Katan M, Cockcroft S. Phosphatidylinositol(4,5)bisphosphate: diverse functions at the plasma membrane. Essays Biochem. 2020;64(3):513–31.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  39. Verma K, Jaiswal R, Paliwal S, Dwivedi J, Sharma S. An insight into PI3k/Akt pathway and associated protein-protein interactions in metabolic syndrome: A recent update. J Cell Biochem. 2023;124(7):923–42.

    Article  CAS  PubMed  Google Scholar 

  40. Ratke J, Entschladen F, Niggemann B, Zänker KS, Lang K. Leptin stimulates the migration of colon carcinoma cells by multiple signaling pathways. Endocrine-related Cancer. 2010;17(1):179–89.

    Article  CAS  PubMed  Google Scholar 

  41. Jiang W, Guo M, Gong M, Chen L, Bi Y, Zhang Y, Shi Y, Qu P, Liu Y, Chen J, et al. Vitamin A bio-modulates apoptosis via the mitochondrial pathway after hypoxic-ischemic brain damage. Mol Brain. 2018;11(1):14.

    Article  PubMed  PubMed Central  Google Scholar 

  42. Uruno A, Sugawara A, Kanatsuka H, Kagechika H, Saito A, Sato K, Kudo M, Takeuchi K, Ito S. Upregulation of nitric oxide production in vascular endothelial cells by all-trans retinoic acid through the phosphoinositide 3-kinase/Akt pathway. Circulation. 2005;112(5):727–36.

    Article  CAS  PubMed  Google Scholar 

  43. Pang X, Yang S, Guo X, Li H, Zhang Y, Wei C, Wang Y, Sun C, Li Y. The association and mediating biomarkers of serum retinol in influencing the development of type 2 diabetes: A prospective cohort study in Middle-Aged and elderly population. Front Nutr. 2022;9:831950.

    Article  PubMed  PubMed Central  Google Scholar 

  44. Majerczyk M, Kocełak P, Choręza P, Arabzada H, Owczarek AJ, Bożentowicz-Wikarek M, Brzozowska A, Szybalska A, Puzianowska-Kuźnicka M, Grodzicki T, et al. Components of metabolic syndrome in relation to plasma levels of retinol binding protein 4 (RBP4) in a cohort of people aged 65 years and older. J Endocrinol Investig. 2018;41(10):1211–9.

    Article  CAS  Google Scholar 

  45. Reay WR, Kiltschewskij DJ, Di Biase MA, Gerring ZF, Kundu K, Surendran P, Greco LA, Clarke ED, Collins CE, Mondul AM, et al. Genetic influences on Circulating retinol and its relationship to human health. Nat Commun. 2024;15(1):1490.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Shin SJ, Chen CH, Kuo WC, Chan HC, Chan HC, Lin KD, Ke LY. Disruption of retinoid homeostasis induces RBP4 overproduction in diabetes: O-GlcNAcylation involved. Metab Clin Exp. 2020;113:154403.

    Article  CAS  PubMed  Google Scholar 

  47. Tian J, Du Y, Wang B, Ji M, Li H, Xia Y, Zhang K, Li Z, Xie W, Gong W et al. Hif1α/Dhrs3a pathway participates in lipid droplet accumulation via retinol and Ppar-γ in fish hepatocytes. Int J Mol Sci 2023, 24(12).

  48. Neuvonen M, Manna M, Mokkila S, Javanainen M, Rog T, Liu Z, Bittman R, Vattulainen I, Ikonen E. Enzymatic oxidation of cholesterol: properties and functional effects of cholestenone in cell membranes. PLoS ONE. 2014;9(8):e103743.

    Article  PubMed  PubMed Central  Google Scholar 

  49. Elia J, Carbonnelle D, Logé C, Ory L, Huvelin JM, Tannoury M, Diab-Assaf M, Petit K, Nazih H. 4-cholesten-3-one decreases breast cancer cell viability and alters membrane raft-localized EGFR expression by reducing lipogenesis and enhancing LXR-dependent cholesterol transporters. Lipids Health Dis. 2019;18(1):168.

    Article  PubMed  PubMed Central  Google Scholar 

  50. Liu T, Ma M, Li K, Tan W, Yu H, Wang L. Biomarkers for bile acid malabsorption in Diarrhea-predominant irritable bowel syndrome: A systematic review and Meta-analysis. J Clin Gastroenterol. 2023;57(5):451–8.

    Article  CAS  PubMed  Google Scholar 

  51. Delle Bovi RJ, Kim J, Suresh P, London E, Miller WT. Sterol structure dependence of insulin receptor and insulin-like growth factor 1 receptor activation. Biochim Et Biophys Acta Biomembr. 2019;1861(4):819–26.

    Article  CAS  Google Scholar 

  52. Pang Y, Kartsonaki C, Lv J, Millwood IY, Fairhurst-Hunter Z, Turnbull I, Bragg F, Hill MR, Yu C, Guo Y, et al. Adiposity, metabolomic biomarkers, and risk of nonalcoholic fatty liver disease: a case-cohort study. Am J Clin Nutr. 2022;115(3):799–810.

    Article  PubMed  Google Scholar 

  53. Aichler M, Borgmann D, Krumsiek J, Buck A, MacDonald PE, Fox JEM, Lyon J, Light PE, Keipert S, Jastroch M, et al. N-acyl taurines and acylcarnitines cause an imbalance in insulin synthesis and secretion provoking Β cell dysfunction in type 2 diabetes. Cell Metabol. 2017;25(6):1334–e13471334.

    Article  CAS  Google Scholar 

  54. Nagao K, Inoue N, Suzuki K, Shimizu T, Yanagita T. The cholesterol metabolite Cholest-5-en-3-One alleviates hyperglycemia and hyperinsulinemia in obese (db/db) mice. Metabolites 2021, 12(1).

  55. Kendel M, Wielgosz-Collin G, Bertrand S, Roussakis C, Bourgougnon N, Bedoux G. Lipid composition, fatty acids and sterols in the seaweeds Ulva Armoricana, and Solieria chordalis from Brittany (France): an analysis from nutritional, chemotaxonomic, and antiproliferative activity perspectives. Mar Drugs. 2015;13(9):5606–28.

    Article  PubMed  PubMed Central  Google Scholar 

  56. Cifuente JO, Comino N, Trastoy B, D’Angelo C, Guerin ME. Structural basis of glycogen metabolism in bacteria. Biochem J. 2019;476(14):2059–92.

    Article  CAS  PubMed  Google Scholar 

  57. van der Pol A, Gil A, Silljé HHW, Tromp J, Ovchinnikova ES, Vreeswijk-Baudoin I, Hoes M, Domian IJ, van de Sluis B, van Deursen JM et al. Accumulation of 5-oxoproline in myocardial dysfunction and the protective effects of OPLAH. Sci Transl Med 2017, 9(415).

  58. Anderson R, Theron AJ, Steel HC, Nel JG, Tintinger GR. ADP-Mediated Upregulation of Expression of CD62P on Human Platelets Is Critically Dependent on Co-Activation of P2Y1 and P2Y12 Receptors. Pharmaceuticals (Basel, Switzerland) 2020, 13(12).

  59. Sun J, Guo F, Ran J, Wu H, Li Y, Wang M, Wang X. Bibliometric and visual analysis of global research on taurine, creatine, carnosine, and Anserine with metabolic syndrome: from 1992 to 2022. Nutrients 2023, 15(15).

  60. Zhang C, Wang Y, Huang F, Zhang Y, Liu Y, Wang Q, Zhang X, Li B, Angwa L, Jiang Y, et al. Fluoride induced metabolic disorder of endothelial cells. Toxicology. 2023;492:153530.

    Article  CAS  PubMed  Google Scholar 

Download references

Acknowledgements

Our gratitude extends to the Department of Cardiovascular I, The First Affiliated Hospital of Heilongjiang University of Traditional Chinese Medicine, for their invaluable support.

Funding

This research was supported by the Graduate Innovative Research Program of Heilongjiang University of Traditional Chinese Medicine (Project No. 2024yjscx007), the Scientific Research Project of the Administration of Traditional Chinese Medicine of Heilongjiang Province (Project No. ZHY2022-148), and the National Natural Science Foundation of China (Grant Nos. 8207150463, 8227152422, 8220153214).

Author information

Authors and Affiliations

  1. Heilongjiang University of Chinese Medicine;, Heilongjiang, P.R. China

    YueGuang Yang, YanLing Ma & Ming Li

  2. The First Department of Cardiovascular, First Affiliated Hospital, Heilongjiang University of Chinese Medicine, 26 Heping Road, Xiangfang, Harbin, Heilongjiang, 150040, P.R. China

    YuBo Han & Li Liu

Authors
  1. YueGuang Yang
    View author publications

    Search author on:PubMed Google Scholar

  2. YanLing Ma
    View author publications

    Search author on:PubMed Google Scholar

  3. Ming Li
    View author publications

    Search author on:PubMed Google Scholar

  4. YuBo Han
    View author publications

    Search author on:PubMed Google Scholar

  5. Li Liu
    View author publications

    Search author on:PubMed Google Scholar

Contributions

YYG was involved in data download and analysis, as well as manuscript writing and editing. MYL and LM contributed to writing the first draft and creating the images. HYB and LL played key roles in manuscript editing, verification, and program/project development. All authors have reviewed and approved the final manuscript.

Corresponding authors

Correspondence to YuBo Han or Li Liu.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, 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 you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. 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-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Yang, Y., Ma, Y., Li, M. et al. Unraveling the causal pathway between phosphatidylinositol, metabolites, and metabolic syndrome: a Mendelian randomization study. Diabetol Metab Syndr 17, 162 (2025). https://doi.org/10.1186/s13098-025-01731-7

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13098-025-01731-7

Keywords

Diabetology & Metabolic Syndrome

ISSN: 1758-5996

Contact us