Skip to main content
Download PDF

Causal relationships between GLP1 receptor agonists, blood lipids, and heart failure: a drug-target mendelian randomization and mediation analysis

Diabetology & Metabolic Syndrome volume 16, Article number: 208 (2024) Cite this article

Abstract

Background

Glucagon-like peptide-1 receptor (GLP1R) agonists have been shown to reduce major cardiovascular events in diabetic patients, but their role in heart failure (HF) remains controversial. Recent evidence implies their potential benefits on cardiometabolism such as lipid metabolism, which may contribute to lowering the risk of HF. Consequently, we designed a Mendelian randomization (MR) study to investigate the causal relationships of circulating lipids mediating GLP1R agonists in HF.

Methods

The available cis-eQTLs for GLP1R target gene were selected as instrumental variables (IVs) of GLP1R agonism. Positive control analyses of type 2 diabetes mellitus (T2DM) and body mass index (BMI) were conducted to validate the enrolled IVs. Two-sample MR was performed to evaluate the associations between GLP1R agonism and HF as well as left ventricular ejection fraction (LVEF). Summary data for HF and LVEF were obtained from two genome-wide association studies (GWASs), which included 977,323 and 40,000 individuals of European ancestry, respectively. The primary method employed was the random-effects inverse variance weighted, with several other methods used for sensitivity analyses, including MR-Egger, MR PRESSO, and weighted median. Additionally, multivariable MR and mediation MR were applied to identify potentially causal lipid as mediator.

Results

A total of 18 independent IVs were included. The positive control analyses showed that GLP1R agonism significantly reduced the risk of T2DM (OR = 0.79, 95% CI = 0.75–0.85, p < 0.0001) and decreased BMI (OR = 0.95, 95% CI = 0.93–0.96, p < 0.0001), ensuring the effectiveness of selected IVs. We found favorable evidence to support the protective effect of GLP1R agonism on HF (OR = 0.75, 95% CI = 0.71–0.79, p < 0.0001), but there was no obvious correlation with increased LVEF (OR = 1.01, 95% CI = 0.95–1.06, p = 0.8332). Among the six blood lipids, only low-density lipoprotein cholesterol (LDL-C) was both associated with GLP1R agonism and HF. The causal effect of GLP1R agonism on HF was partially mediated through LDL-C by 4.23% of the total effect (95% CI = 1.04–7.42%, p = 0.0093).

Conclusions

This study supported the causal relationships of GLP1R agonists with a reduced risk of HF. LDL-C might be the mediator in this association, highlighting the cardiometabolic benefit of GLP1R agonists on HF.

Graphical Abstract

Introduction

Glucagon-like peptide-1 receptor (GLP1R) agonists represent a novel antidiabetic medications that activate the receptor of the gut-derived hormone GLP1 to increase insulin secretion and inhibit glucagon secretion, thereby attaining glycaemic control [1]. Besides, several GLP1R agonists have been approved by the FDA to lower body weight in obese individuals [2]. Notably, growing evidence has shown the benefits of reducing the risk of major cardiovascular events (MACE), a composite of cardiovascular death, nonfatal myocardial infarction, and nonfatal stroke, in individuals with type 2 diabetes mellitus (T2DM) [3, 4]. However, unlike sodium-glucose co-transporter 2 (SGLT2) inhibitors [5], the current evidence of GLP1R agonists in the prevention and progression of heart failure (HF) is controversial, and the potential cardiometabolic regulatory mechanism remains unclear [6, 7].

Abnormalities in lipid metabolism has been associated with an elevated risk of HF [8, 9]. GLP1R agonists have exhibited favorable effects on the improvement of cardiometabolic dysregulation, particularly lipid metabolism in recent clinical studies and animal experiments [10,11,12]. Nevertheless, certain studies failed to observe significant changes in routinely measured plasma lipid levels [13]. The discrepancy in these studies might be attributed in part to the limited sample size and the existence of residual confounding factors. Therefore, the causal effect of GLP1R agonists on HF and whether blood lipids mediate this relationship are still underexplored.

Mendelian randomization (MR) utilizes genetic variations related to exposure as instrumental variables (IVs), which has become a commonly accepted approach to assess potential causal relationships between different traits [14]. MR imitates the typical randomization process of randomized controlled trials by randomly allocating genetic variants during gametogenesis, thereby controlling potential confounders and avoiding reverse causation bias [15].

In light of these premises, we conducted a drug-target and mediation MR analysis to explore the causal relationship between GLP1R agonists and HF and the potential role of blood lipid as mediator, which will help clarify the underlying cardiometabolic benefit of GLP1R agonists on HF.

Methods

Study design

This study conformed to the STROBE-MR statement [16]. MR analysis relied on three major assumptions: (1) the relevance assumption, which states that IVs are strongly connected with GLP1R; (2) the independence assumption, which stipulates that IVs are independent of confounders; and (3) the exclusion restriction assumption, which states that IVs affect outcomes only through exposure.

The overall of our study design is displayed in Fig. 1 as four steps. We first selected genetic instruments for GLP1R and performed positive control analyses between GLP1R agonism and T2DM, as well as body mass index (BMI). Next, we carried out a drug-targeted MR to investigate the causal relationships of GLP1R agonism with HF and cardiac function. Then, we identified and prioritized causal lipids for HF by multivariable MR (MVMR). Lastly, we evaluated the interactions from GLP1R agonism to HF via lipid mediator.

Fig. 1
figure 1

Study design

Full size image

Data sources

All used genome-wide association studies (GWASs) data on T2DM, BMI, blood lipids, cardiac function, and HF in our study are publicly available, as listed in Table S1.

T2DM and BMI were employed for the positive control analysis. The GWAS data for T2DM originated from a large GWAS meta-analysis containing 61,714 T2DM cases and 593,952 controls of European ancestry, by combining 3 GWASs: Diabetes Genetics Replication and Meta-analysis (DIAGRAM), Genetic Epidemiology Research on Aging (GERA), and the full cohort release from the UK Biobank (UKB) [17]. Summary statistics for BMI was derived from a large-scale GWAS conducted by Genetic Investigation of ANthropometric Traits (GIANT) involving 681,275 subjects of European ancestry [18].

Blood lipids routinely measured in the clinic were used as the mediation. Data for triglycerides (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C) and low-density lipoprotein cholesterol (LDL-C) were collected from the Global Lipid Genetics Consortium (GLGC) [19]. We extracted GWAS data involving 94,595 individuals of European ancestry for this study. Moreover, we obtained genetic associations with apolipoprotein A-1 (ApoA1) (sample size = 393,193) and apolipoprotein B (ApoB) (sample size = 439,214) from the UKB [20].

HF was the primary outcome in our analysis. The summary level data on HF for primary analysis was obtained from an European ancestry GWAS meta-analysis involving 29 datasets, which comprised 47,309 HF cases and 930,014 controls [21]. All types of HF were included, and the diagnosis of HF was confirmed according to hospitalization, physician diagnosis or death records.

HF is usually associated with impairment of left ventricular function manifesting with either reduced or preserved ejection fraction. Thus, the left ventricular ejection fraction (LVEF) was further analyzed as the secondary outcome. Data for cardiac function originated from a GWAS of 40,000 individuals in the UKB [22]. The genetic variations in LVEF were obtained based on cardiovascular magnetic resonance imaging (CMRI).

Genetic instruments for GLP1R agonism

As described in a recent investigation [23], the selection of genetic variants for GLP1R agonism followed these steps. We obtained cis-eQTL regions within ± 100 kb of the GLP1R gene (GRCh37.p13: chromosome 6: 39016557, 39059079) from the eQTLGen Consortium [24]. The common variants significantly (p < 5.0 × 10− 8) associated with the expression of GLP1R in blood and possessing an effect allele frequency (EAF) exceeding 1% were selected as the available genetic proxies. In order to avoid the effect of strong linkage disequilibrium (LD), the threshold of r2 < 0.001 was set. Moreover, the LD trait tool (https://ldlink.nci.nih.gov/) was utilized to further exclude single nucleotide polymorphisms (SNPs) directly related to the outcomes (R2 < 0.30).

After selecting IVs of GLP1R agonism, positive controls were conducted to validate the association of genetic instruments of GLP1R agonism with endpoints, including T2DM and BMI.

Genetic instruments for lipids and HF

During the genetic instruments for blood lipids and HF, we selected SNPs with a genome-wide significance p threshold (< 5.0 × 10− 8). Then we clumped those SNPs to an LD of r2 < 0.0001 and clump distance = 10,000 kb using the European reference panel of the 1000 Genomes Project. If there were palindromic SNPs, the allele frequency data was used to identify the forward strand allele. For the SNPs that were unfound in the outcome, we searched for proxies with r2 ≥ 0.8 as substitutes. Finally, the F-statistics of all SNPs were calculated to exclude the weak instrumental variables.

Statistical analysis

The random-effect inverse variance weighted (IVW) method was employed as the primary approach. This method combines the Wald ratio for individual SNPs and provides the most powerful estimates [25]. To account for multiple hypothesis testing, the Benjamini-Hochberg false discovery rate (FDR) was applied [26]. The FDR-corrected p value < 0.10 was considered significant [27]. To investigate the heterogeneity of IVW, the Cochran’s Q test was utilized. Notably, the IVW method could still ensure the robustness and conservativeness of results in the case of heterogeneity existing (p value < 0.05) [28]. Additionally, the sensitivity analyses were further performed by using a series of methods, including MR-Egger, MR pleiotropy residual sum and outlier (MR PRESSO), weighted median, simple mode, and weighted mode. The intercept of MR-Egger regression was applied to examine pleiotropy and p value < 0.05 indicates the presence of horizontal pleiotropy [29]. MR PRESSO was used to check for possible outliers in IVW and correct MR estimates after removing outliers if it exists [30]. Finally, the leave-one-out method was utilized to examine whether the MR results were reliable.

In order to estimate the mediating role of blood lipid in the associations between GLP1R agonism and HF, we conducted a two-step MR analysis as follows. First, we evaluated the effect of GLP1R agonism on blood lipids using two-sample MR (β1). Second, MVMR was performed for estimating the effects of those blood lipids showing a significant association with GLP1R agonism on HF (β2). The mediated proportion of blood lipid as mediator was calculated as the indirect effect (β1 × β2) divided by the total effect (β0), while the 95% confidence interval (CI) of mediation effect was calculated by the delta method [31]. In MVMR, we used MV-IVW as the primary result in case of without pleiotropy. Besides, two other methods were employed to validate the robustness, including the MVMR-Egger and MVMR-Lasso [32].

The results of the cause effect are presented as odds ratio (OR) and 95% CI. All MR analyses were executed in the program R (v.4.3.3) using packages “TwoSampleMR”, “MRPRESSO”, “MendelianRandomization”, “MVMR”, and “forestploter”.

Results

Positive control analysis

A total of 18 independent SNPs were finally chosen from eQTLGen as genetic instruments for GLP1R agonism, all of which had an F statistics > 20 (Table S2), suggesting no weak instrumental bias. As expected, IVW results demonstrated that GLP1R agonism significantly reduced the risk of T2DM (OR = 0.79, 95% CI = 0.75–0.85, p < 0.0001) and obviously lowered BMI (OR = 0.95, 95% CI = 0.93–0.96, p < 0.0001) (Table 1). These results were supported by the sensitivity analyses of MR PRESSO and weighted median. All p values were greater than 0.05 in the test of heterogeneity and pleiotropy, which implied no evidence of heterogeneity or horizontal pleiotropy. The positive control analyses illustrated the credible association between genetic instruments of GLP1R agonism and T2DM as well as BMI, suggesting the effectiveness of IVs.

Table 1 Positive control analyses of GLP1R agonism
Full size table

Causal associations of GLP1R agonism with HF and LVEF

Genetically predicted GLP1R agonism had an obvious protective effect on HF (OR = 0.75, 95% CI = 0.71–0.79, p < 0.0001), while no significant association was found between genetically determined GLP1R agonism and LVEF (OR = 1.01, 95% CI = 0.95–1.06, p = 0.8332) (Fig. 2). The heterogeneity was not observed (Q = 4.485, p = 0.992; Q = 10.835, p = 0.699), and there was also no directional pleiotropy (Egger intercept=-0.016, p = 0.109; Egger intercept=-0.009, p = 0.302) (Table S3).

Fig. 2
figure 2

The forest plot of showing the effects of GLP1R agonism on heart failure and left ventricular ejection fraction. NSNP, number of single nucleotide polymorphism; Q_P, p value for heterogeneity of Q statistics; Intercept (P), p value for intercept of MR-Egger regression

Full size image

Mediation MR of GLP1R agonism, blood lipids, and HF

We examined the effects of GLP1R agonism on blood lipids and identified two lipids potentially associated with GLP1R agonism (Fig. 3, Table S4). The most significant result was for LDL-C (OR = 0.93, 95% CI = 0.89–0.98, p = 0.005, FDR = 0.03). The sensitivity analyses of MR PRESSO and weighted median further supported such causal relationship. Another significant lipid was TC (OR = 0.95, 95% CI = 0.91–0.99, p = 0.0275, FDR = 0.0825), with no evidence of heterogeneity (Q = 1.358, p = 0.999) and pleiotropy (Egger intercept=-0.005, p = 0.493).

Fig. 3
figure 3

The forest plot of showing the effects of GLP1R agonism on blood lipids. NSNP, number of single nucleotide polymorphism; Q_P, p value for heterogeneity of Q statistics; Intercept (P), p value for intercept of MR-Egger regression

Full size image

We further estimated the causal relationships of blood lipids on HF. All genetic instruments had sufficient strength for F statistics greater than 10 (Table S5). We found that HDL-C had a negative association with HF (OR = 0.91, 95% CI = 0.84–0.99, p = 0.0311, FDR = 0.0373). The results also showed that TG (OR = 1.14, 95% CI = 1.06–1.22, p = 0.0002, FDR = 0.0005), TC (OR = 1.12, 95% CI = 1.05–1.20, p = 0.0004, FDR = 0.0008), LDL-C (OR = 1.14, 95% CI = 1.07–1.21, p < 0.0001, FDR = 0.0004) and ApoB (OR = 1.16, 95% CI = 1.05–1.27, p = 0.0022, FDR = 0.0033) had causal effects on HF. The MR-Egger regression intercept analysis suggested no evidence of horizontal pleiotropy. Despite the heterogeneity test for IVW showed significant p values (< 0.05), indicating potential heterogeneity, the MR PRESSO results were generally consistent with IVW after removing outliers, which revealed that underlying heterogeneity did not greatly bias our results (Table S6). Besides, leave-one-out analyses demonstrated that the overall effects were unlikely to be violated by certain extreme SNPs, indicating the robustness of results (Fig. 4).

Fig. 4
figure 4

Leave-one-out analysis showing the effects of each exposure on heart failure. (A) Triglycerides; (B) Total cholesterol; (C) Low-density lipoprotein cholesterol; (D) High-density lipoprotein cholesterol; (E) Apolipoprotein A-1; (F) Apolipoprotein B

Full size image

However, the results of reverse-direction MR analysis showed no potential effects of HF on six blood lipids (Table S7).

In the MVMR, we included TC and LDL-C which were significantly linked to both GLP1R agonism and HF. The MV-IVW results exerted strong evidence for a positive causal effect between LDL-C and the risk of HF (OR = 1.19, 95% CI = 1.02–1.39, p = 0.0297), but no evidence between TC and HF (Table 2). The consistent findings were also observed using MVMR-Egger and MVMR-Lasso methods. Moreover, the MVMR-Egger intercept test revealed no horizontal pleiotropy (Egger intercept=-0.001, p = 0.817).

Table 2 Multivariable MR estimates of the causal effect of blood lipids on heart failure
Full size table

Finally, we reported a causal effect to highlight the mediation role of LDL-C in the relationship between GLP1R agonism and HF, with a mediated proportion of 4.23% (95% CI = 1.04-7.42%, p = 0.0093) (Fig. 5).

Fig. 5
figure 5

The LDL-C mediated the causal effect of GLP1R agonists on heart failure

Full size image

Discussion

In the present investigation, we employed human genetic data to identify proxies for GLP1R agonism. We then performed univariable, multivariable, and mediation MR analyses to evaluate the causal associations between GLP1R agonism and HF, and the role of blood lipids as mediators. Our results supported the protective effect of GLP1R agonists on the risk of HF, but found no association with increased LVEF. Moreover, LDL-C may mediate 4.23% of the causal effect of GLP1R agonism on HF.

In recent years, research on GLP1R agonists for cardiovascular and renal outcomes has been burst. Despite several large clinical trials investigating the role of GLP1R agonists in HF, these observations are controversial and “hard” results widely accepted within the scientific community are still lacking [33]. The recently published STEP-HFpEF trial revealed that semaglutide produced greater improvements in HF-related symptoms, physical limitations, and exercise function [34]. Several meta-analyses also suggested that GLP1R agonists have moderate benefits on the reduction of HF events in patients with T2DM [3, 35]. However, there are other clinical trials with inconsistent results. In the LEADER study, liraglutide showed a non-significant 13% reduction in the risk of HF hospitalization (HR = 0.87, 95% CI = 0.73–1.05) [36]. A similar neutral result of dulaglutide on HF events was also observed (HR = 0.93, 95% CI = 0.77–1.12) [37]. The latest SELECT trial reported that semaglutide led to a 20% reduction in MACE in obesity without diabetes, but failed to show a positive effect on cardiovascular death. Hence, the investigators did not report between-group differences for secondary endpoints, including HF [4]. Apart from HF, our analysis also encompassed the clinical indicator of cardiac function and demonstrated no genetic evidence supporting the benefit of GLP1R agonism on LVEF, aligning with the clinical results from the LIVE study [38].

Until recently, none of the trials included HF events as a component of the primary outcome, but rather as a prespecified secondary outcome, and few studies were conducted on the characterization of HF in terms of LVEF or biomarker levels [6]. On the other hand, previous trials generally focused on patients with T2DM. While the relevance of HF with T2DM is undisputed, it is unclear whether this positive correlation of GLP1R agonists can be extended to the general population. In addition, given the complexity of comorbidities and concomitant medications in patients with clinical HF, the trials may be potentially influenced by kinds of confounding factors. To address these challenges, we designed an MR study to provide valuable evidence for the protective effect of GLP1R agonists on HF.

The potent cardiovascular benefits of GLP1R agonists may partially result from glycaemic control and body weight loss, and it has also been suggested to be associated with improvements in lipids [39]. Nevertheless, previous studies about the effects of GLP1R agonists on blood lipids have yielded inconsistent results. A single-center randomized controlled study observed a significant decrease only in TG levels after 16 weeks of liraglutide treatment in T2DM patients [40]. Another trial by Peradez et al. showed that treatment with high-dose liraglutide induced changes in serum lipid profiles of obese patients at five weeks, with TC and free cholesterol being the most significant whereas HDL-C showing only minor changes [41]. A clinical trial on subjects with obesity performed by Hjerpsted et al. implied that fasting TC, HDL-C and TG were lower with semaglutide, but no difference was observed for LDL-C and ApoB compared with placebo [42]. Moreover, several studies have demonstrated that GLP1R agonists fail to show an effect in improving lipid profiles routinely measured [13, 43]. Of note, the aforementioned trials were carried out in the presence of statin use, which makes it difficult to define the lipid-lowering effects of GLP1R agonists. In our MR study, we found that GLP1R agonism could significantly decrease TC and LDL-C levels, offering genetic evidence of GLP1R agonists in the improvement of blood lipids.

It has been well documented that dyslipidaemias are associated with the risk of atherosclerosis diseases such as coronary artery disease, a well-known cause of ischaemic heart failure [44]. However, research on the relationships between circulating lipids and HF has been less investigated. An observational study suggested that LDL-C was unlikely to be the risk factor of incident HF [45]. One previous MR study identified a positive association of genetically predicted LDL-C with HF [46], in line with our results. Therefore, LDL-C might be the mediator in the causal association between GLP1R agonism and HF. Our study provided genetic evidence for the cardiometabolic benefit of GLP1R agonists on HF.

This study represents the first to employ an MR design to investigate the causal associations among GLP1R agonists, blood lipids, and HF in the general population. Meanwhile, limitations also should to be noticed. Firstly, MR findings reflect the long-period effects of genetic proxies of GLP1R agonism on blood lipids and HF, which may not fully align with results observed in clinical trials over a relatively limited time. However, the expected causality could provide a direction for clinical trials and animal experiments. Secondly, there was approximately 40% overlapping participants between the GWAS datasets of blood lipids and HF, indicating the possibility of bias. In the case of the exposure and outcome data originating from overlapping samples, this overlap may bias the estimated causal effect by introducing an association between IVs and confounders. If combining with the weak instruments, the Type 1 error will be inflated [47]. However, the F statistics of IVs were sufficiently high (all > 27). Meanwhile, it was also unlikely that sample overlap would bias our results, as the estimated bias was negligible (bias = 0.002) and the Type 1 error rate was low using a website-based tool (https://sb452.shinyapps.io/overlap/). Thirdly, the pathophysiology of distinct HF subtypes, such as HFpEF and HFrEF, may differ slightly. Due to the limited availability of GWAS specific to certain types of HF, our findings were challenging to distinguish the potential effects of GLP1R agonists on single HF subtype. Lastly, our study was performed exclusively on European populations, so it should be caution to generalize these findings to other ancestries.

Conclusions

In summary, our study supported the causal relationships between GLP1R agonists, blood lipids, and HF. Specifically, LDL-C appears to mediate the effect of GLP1R agonism on HF. In addition, we found that GLP1R agonism was no association with increased LVEF. This current study provides genetic evidence with clinical applications for GLP1R agonists, highlighting the cardiometabolic benefit of GLP1R agonists on HF.

Data availability

The GWAS Summary statistics used in this study were publicly accessed from the IEU OpenGWAS project (https://gwas.mrcieu.ac.uk/), GWAS Catalog (https://www.ebi.ac.uk/gwas/home), the eQTLGen Consortium (https://eqtlgen.org/) and Heart-KP (http://heartkp.org/).

Abbreviations

ApoA1:

Apolipoprotein A-1

ApoB:

Apolipoprotein B

BMI:

Body mass index

CI:

Confidence interval

FDR:

False discovery rate

GLP1R:

Glucagon-like peptide-1 receptor

HDL-C:

High-density lipoprotein cholesterol

HF:

Heart failure

HR:

Hazard ratio

IVs:

Instrumental variables

IVW:

Inverse variance weighted

LDL-C:

Low-density lipoprotein cholesterol

LVEF:

Left ventricular ejection fraction

MACE:

Major cardiovascular events

MR:

Mendelian randomization

MVMR:

Multivariable Mendelian randomization

OR:

Odds ratio

SNP:

Single nucleotide polymorphism

T2DM:

Type 2 diabetes mellitus

TC:

Total cholesterol

TG:

Triglycerides

References

  1. Andersen A, Knop FK, Vilsbøll T. A pharmacological and clinical overview of oral semaglutide for the treatment of type 2 diabetes. Drugs. 2021;81(9):1003–30.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  2. Haddad F, Dokmak G, Bader M, Karaman R. A Comprehensive Review on Weight loss Associated with anti-diabetic medications. Life (Basel). 2023;13(4).

  3. Sattar N, Lee MMY, Kristensen SL, et al. Cardiovascular, mortality, and kidney outcomes with GLP-1 receptor agonists in patients with type 2 diabetes: a systematic review and meta-analysis of randomised trials. Lancet Diabetes Endocrinol. 2021;9(10):653–62.

    Article  CAS  PubMed  Google Scholar 

  4. Lincoff AM, Brown-Frandsen K, Colhoun HM, et al. Semaglutide and Cardiovascular outcomes in obesity without diabetes. N Engl J Med. 2023;389(24):2221–32.

    Article  CAS  PubMed  Google Scholar 

  5. Ali MU, Mancini GBJ, Fitzpatrick-Lewis D, et al. The effectiveness of sodium-glucose co-transporter 2 inhibitors on cardiorenal outcomes: an updated systematic review and meta-analysis. Cardiovasc Diabetol. 2024;23(1):72.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Seferović PM, Coats AJS, Ponikowski P, et al. European Society of Cardiology/Heart Failure Association position paper on the role and safety of new glucose-lowering drugs in patients with heart failure. Eur J Heart Fail. 2020;22(2):196–213.

    Article  PubMed  Google Scholar 

  7. Villaschi A, Ferrante G, Cannata F, et al. GLP-1-ra and heart failure-related outcomes in patients with and without history of heart failure: an updated systematic review and meta-analysis. Clin Res Cardiol. 2024;113(6):898–909.

    Article  PubMed  Google Scholar 

  8. Varbo A, Nordestgaard BG. Nonfasting triglycerides, Low-Density Lipoprotein Cholesterol, and heart failure risk: two cohort studies of 113 554 individuals. Arterioscler Thromb Vasc Biol. 2018;38(2):464–72.

    Article  CAS  PubMed  Google Scholar 

  9. Hage C, Löfgren L, Michopoulos F, et al. Metabolomic Profile in HFpEF vs HFrEF patients. J Card Fail. 2020;26(12):1050–9.

    Article  PubMed  Google Scholar 

  10. Yao H, Zhang A, Li D, et al. Comparative effectiveness of GLP-1 receptor agonists on glycaemic control, body weight, and lipid profile for type 2 diabetes: systematic review and network meta-analysis. BMJ. 2024;384:e076410.

    Article  PubMed  PubMed Central  Google Scholar 

  11. Kosiborod MN, Bhatta M, Davies M, et al. Semaglutide improves cardiometabolic risk factors in adults with overweight or obesity: STEP 1 and 4 exploratory analyses. Diabetes Obes Metab. 2023;25(2):468–78.

    Article  CAS  PubMed  Google Scholar 

  12. Gallo G, Volpe M. Potential mechanisms of the Protective effects of the Cardiometabolic drugs Type-2 sodium-glucose transporter inhibitors and glucagon-like Peptide-1 receptor agonists in Heart failure. Int J Mol Sci 2024;25(5).

  13. Jendle J, Hyötyläinen T, Orešič M, Nyström T. Pharmacometabolomic profiles in type 2 diabetic subjects treated with liraglutide or glimepiride. Cardiovasc Diabetol. 2021;20(1):237.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Emdin CA, Khera AV, Kathiresan S, Mendelian Randomization. JAMA. 2017;318(19):1925–6.

    Article  PubMed  Google Scholar 

  15. Larsson SC, Butterworth AS, Burgess S. Mendelian randomization for cardiovascular diseases: principles and applications. Eur Heart J. 2023;44(47):4913–24.

    Article  PubMed  PubMed Central  Google Scholar 

  16. Skrivankova VW, Richmond RC, Woolf BAR, 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 

  17. Xue A, Wu Y, Zhu Z, et al. Genome-wide association analyses identify 143 risk variants and putative regulatory mechanisms for type 2 diabetes. Nat Commun. 2018;9(1):2941.

    Article  PubMed  PubMed Central  Google Scholar 

  18. Yengo L, Sidorenko J, Kemper KE, et al. Meta-analysis of genome-wide association studies for height and body mass index in ∼700000 individuals of European ancestry. Hum Mol Genet. 2018;27(20):3641–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Willer CJ, Schmidt EM, Sengupta S, et al. Discovery and refinement of loci associated with lipid levels. Nat Genet. 2013;45(11):1274–83.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. Richardson TG, Sanderson E, Palmer TM, et al. 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.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Shah S, Henry A, Roselli C, et al. Genome-wide association and mendelian randomisation analysis provide insights into the pathogenesis of heart failure. Nat Commun. 2020;11(1):163.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Zhao B, Li T, Fan Z, et al. Heart-brain connections: phenotypic and genetic insights from magnetic resonance images. Science. 2023;380(6648):abn6598.

    Article  CAS  PubMed  Google Scholar 

  23. Sun Y, Liu Y, Dian Y, Zeng F, Deng G, Lei S. Association of glucagon-like peptide-1 receptor agonists with risk of cancers-evidence from a drug target mendelian randomization and clinical trials. Int J Surg. 2024.

  24. Võsa U, Claringbould A, Westra H-J, et al. Large-scale cis- and trans-eQTL analyses identify thousands of genetic loci and polygenic scores that regulate blood gene expression. Nat Genet. 2021;53(9):1300–10.

    Article  PubMed  PubMed Central  Google Scholar 

  25. Stephen B, George DS, Neil MD, et al. Guidelines for performing mendelian randomization investigations: update for summer 2023. Wellcome Open Res. 2023;4:186.

    Article  Google Scholar 

  26. Benjamini Y, Drai D, Elmer G, Kafkafi N, Golani I. Controlling the false discovery rate in behavior genetics research. Behav Brain Res. 2001;125(1–2):279–84.

    Article  CAS  PubMed  Google Scholar 

  27. Guo W, Zhao L, Huang W, et al. Sodium-glucose cotransporter 2 inhibitors, inflammation, and heart failure: a two-sample mendelian randomization study. Cardiovasc Diabetol. 2024;23(1):118.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Burgess S, Small DS, Thompson SG. A review of instrumental variable estimators for mendelian randomization. Stat Methods Med Res. 2017;26(5):2333–55.

    Article  PubMed  Google Scholar 

  29. Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. 2015;44(2):512–25.

    Article  PubMed  PubMed Central  Google Scholar 

  30. Verbanck M, Chen C-Y, 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. Li J, Yu Y, Sun Y, et al. SGLT2 inhibition, circulating metabolites, and atrial fibrillation: a mendelian randomization study. Cardiovasc Diabetol. 2023;22(1):278.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Sun W, Wu X, Yang H, et al. Identifying causal associations between women’s reproductive traits and risk of schizophrenia: a multivariate validated two-sample mendelian randomization analysis. BMC Psychiatry. 2024;24(1):161.

    Article  PubMed  PubMed Central  Google Scholar 

  33. Khan MS, Fonarow GC, McGuire DK, et al. Glucagon-like peptide 1 receptor agonists and heart failure: the need for further evidence generation and practice guidelines optimization. Circulation. 2020;142(12):1205–18.

    Article  CAS  PubMed  Google Scholar 

  34. Kosiborod MN, Abildstrøm SZ, Borlaug BA, et al. Semaglutide in patients with heart failure with preserved ejection fraction and obesity. N Engl J Med. 2023;389(12):1069–84.

    Article  CAS  PubMed  Google Scholar 

  35. Giugliano D, Scappaticcio L, Longo M, et al. GLP-1 receptor agonists and cardiorenal outcomes in type 2 diabetes: an updated meta-analysis of eight CVOTs. Cardiovasc Diabetol. 2021;20(1):189.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Marso SP, Daniels GH, Brown-Frandsen K, et al. Liraglutide and Cardiovascular outcomes in type 2 diabetes. N Engl J Med. 2016;375(4):311–22.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Gerstein HC, Colhoun HM, Dagenais GR, et al. Dulaglutide and cardiovascular outcomes in type 2 diabetes (REWIND): a double-blind, randomised placebo-controlled trial. Lancet. 2019;394(10193):121–30.

    Article  CAS  PubMed  Google Scholar 

  38. Jorsal A, Kistorp C, Holmager P, et al. Effect of liraglutide, a glucagon-like peptide-1 analogue, on left ventricular function in stable chronic heart failure patients with and without diabetes (LIVE)-a multicentre, double-blind, randomised, placebo-controlled trial. Eur J Heart Fail. 2017;19(1):69–77.

    Article  CAS  PubMed  Google Scholar 

  39. Prasad-Reddy L, Isaacs D. A clinical review of GLP-1 receptor agonists: efficacy and safety in diabetes and beyond. Drugs Context. 2015;4:212283.

    Article  PubMed  PubMed Central  Google Scholar 

  40. Matikainen N, Söderlund S, Björnson E, et al. Liraglutide treatment improves postprandial lipid metabolism and cardiometabolic risk factors in humans with adequately controlled type 2 diabetes: a single-centre randomized controlled study. Diabetes Obes Metab. 2019;21(1):84–94.

    Article  CAS  PubMed  Google Scholar 

  41. Peradze N, Farr OM, Perakakis N, Lázaro I, Sala-Vila A, Mantzoros CS. Short-term treatment with high dose liraglutide improves lipid and lipoprotein profile and changes hormonal mediators of lipid metabolism in obese patients with no overt type 2 diabetes mellitus: a randomized, placebo-controlled, cross-over, double-blind clinical trial. Cardiovasc Diabetol. 2019;18(1):141.

    Article  PubMed  PubMed Central  Google Scholar 

  42. Hjerpsted JB, Flint A, Brooks A, Axelsen MB, Kvist T, Blundell J. Semaglutide improves postprandial glucose and lipid metabolism, and delays first-hour gastric emptying in subjects with obesity. Diabetes Obes Metab. 2018;20(3):610–9.

    Article  CAS  PubMed  Google Scholar 

  43. Widiarti W, Sukmajaya AC, Nugraha D, Alkaff FF. Cardioprotective properties of glucagon-like peptide-1 receptor agonists in type 2 diabetes mellitus patients: a systematic review. Diabetes Metab Syndr. 2021;15(3):837–43.

    Article  CAS  PubMed  Google Scholar 

  44. Kemp CD, Conte JV. The pathophysiology of heart failure. Cardiovasc Pathol. 2012;21(5):365–71.

    Article  CAS  PubMed  Google Scholar 

  45. Lind L, Ingelsson M, Sundstrom J, Ärnlöv J. Impact of risk factors for major cardiovascular diseases: a comparison of life-time observational and mendelian randomisation findings. Open Heart 2021;8(2).

  46. Allara E, Morani G, Carter P, et al. Genetic determinants of lipids and Cardiovascular Disease outcomes: a wide-angled mendelian randomization investigation. Circ Genom Precis Med. 2019;12(12):e002711.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  47. Burgess S, Davies NM, Thompson SG. Bias due to participant overlap in two-sample mendelian randomization. Genet Epidemiol. 2016;40(7):597–608.

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We thank the participants in all the GWASs used in this study and the investigators who made these GWAS data publicly available.

Funding

This work was supported by the Chinese Medicine inheritance and innovation “thousand million” Talents Project (Qi Huang Project 2021) Qi Huang Scholars, National Administration of Traditional Chinese Medicine High-level TCM Key discipline Project (grant number: zyyzdxk-2023253), National Natural Science Foundation of China (grant number: 82374407).

Author information

Authors and Affiliations

  1. Dongzhimen Hospital, Beijing University of Chinese Medicine, Beijing, 100700, China

    Tianshi Mao, Long Xie, Xinyan Qu, Ruli Feng, Yi Pan, Qun Gao & Qian Lin

  2. The First Affiliated Hospital of Anhui Medical University, Hefei, 230032, China

    Jie Chen

  3. Dongfang Hospital, Beijing University of Chinese Medicine, Beijing, 100078, China

    Tong Su, Jie Wan & Xiaoyun Cui

  4. Beijing University of Chinese Medicine Third Affiliated Hospital, Beijing, 100029, China

    Wenhao Jia

Authors
  1. Tianshi Mao
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Jie Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Tong Su
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Long Xie
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Xinyan Qu
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Ruli Feng
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Yi Pan
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Jie Wan
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Xiaoyun Cui
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Wenhao Jia
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Qun Gao
    View author publications

    You can also search for this author in PubMed Google Scholar

  12. Qian Lin
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

TS.M, Q.G, and Q.L had full access to all of the data in the study and takes responsibility for the integrity and accuracy of the data analysis. Concept and design: TS.M, Q.G, and Q.L. Statistical analysis: TS.M, J.C, and T.S. Acquisition, analysis or interpretation of data: L.X, XY.Q, RL.F, and Y.P. Preparation of figures and tables: J.W, XY.C, and WH.J. Drafting of the manuscript: TS.M, J.C, and T.S. Revision of the manuscript: All authors. Obtained funding: Q.L. Supervision: Q.G. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Qun Gao or Qian Lin.

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

Mao, T., Chen, J., Su, T. et al. Causal relationships between GLP1 receptor agonists, blood lipids, and heart failure: a drug-target mendelian randomization and mediation analysis. Diabetol Metab Syndr 16, 208 (2024). https://doi.org/10.1186/s13098-024-01448-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13098-024-01448-z

Keywords

Diabetology & Metabolic Syndrome

ISSN: 1758-5996

Contact us