Association of new subtypes of diabetes with cardiovascular disease: Insights from Mendelian randomization

doi:10.21203/rs.3.rs-5230862/v1

Download PDF

Research Article

Association of new subtypes of diabetes with cardiovascular disease: Insights from Mendelian randomization

https://doi.org/10.21203/rs.3.rs-5230862/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Purpose: Globally, the incidence of cardiovascular disease-related mortality has gradually increased, with higher cardiovascular morbidity in individuals with diabetes. In 2018, researchers stratified new diabetes subgroups using cluster analyses; however, it is still unclear whether a causal relationship exists between these diabetes subtypes and cardiovascular disease. Thus, in this study, we aimed to assess the potential causal relationship between the subtypes of diabetes mellitus and cardiovascular disease using two-sample Mendelian randomization analysis.

Methods: Two-sample Mendelian randomization was used to analyze causal associations between the diabetes subtype and cardiovascular disease using inverse variance-weighted analysis, which was the primary analytical method.

Results Severe insulin-deficient diabetes (SIDD) was found to be a potential risk factor for peripheral arterial disease (odds ratio [OR] = 1.03; 95% confidence interval [Cl]=1.00–1.06; P=0.048). Severe autoimmune diabetes (SAID) was a risk factor for peripheral arterial disease (OR=1.03; 95% Cl=1.01–1.05; P=0.004), and a potential protective factor against atrial fibrillation (OR=0.99; 95% Cl=0.98–1.00; P=0.049).

Conclusion: SAID bears a positive causal association with peripheral arterial disease, with evidence suggesting a negative association with atrial fibrillation. SIDD bears a positive causal association with peripheral arterial disease. This study facilitates further enhancement of cardiovascular disease management by clarifying the subtype of diabetes and cardiovascular interrelationships.

Mendelian randomization

diabetes subtypes

cardiovascular disease

severe autoimmune diabetes

severe insulin-deficient diabetes

peripheral arterial disease

atrial fibrillation

In recent years, the incidence of deaths due to cardiovascular diseases has increased worldwide [1]. Although epidemiological studies and clinical trials have identified several risk factors for cardiovascular disease, the etiology is not yet fully understood [2]. Currently, the global disease burden of diabetes is substantial; patients with diabetes are at a higher risk of cardiovascular disease, with cardiovascular complications being an important cause of death [3]. In 2018, a new subgroup stratification of diabetes was developed, which helps provide individualized and precise treatment for people with diabetes. This classification organizes diabetes into five types, viz. severe autoimmune diabetes (SAID), severe insulin-deficient diabetes (SIDD), severe insulin-resistant diabetes (SIRD), mild obesity-related diabetes (MOD), and mild age-related diabetes (MARD) [4]. Several cross-sectional studies and clinical trials have found that these diabetes subtypes are strongly associated with the development of cardiovascular disease [5,6].

However, in-depth exploration of the relationship between the diabetes subtypes and cardiovascular disease has been prevented by several factors, including the inability to determine causal associations and the presence of confounding influences, which are inherent limitations of observational studies; thus, the results should be interpreted cautiously. Additionally, experimental studies are time-consuming and labor-intensive and are associated with have ethical and moral issues.

Mendelian randomization (MR) analyses, which are considered to be natural randomized controlled experiments [7], can assess potential causal relationships between exposure and clinical outcomes using genetic markers. MR is widely used in research because of its ability to infer causality through randomly assigned genetic variants unaffected by individual behavioral or environmental factors, thus avoiding reverse causation bias [8]. Recent MR studies have indicated that compared to type 2 diabetes, type 1 diabetes contributes more significantly to cardiovascular disease [9]. However, diabetes is heterogeneous, with different subtypes posing differential cardiovascular risks [10].

Currently, the extent of association between the diabetes subtype and cardiovascular disease has not been clearly assessed. This study aimed to clarify the causal relationship between diabetes subtypes and cardiovascular disease using two-sample MR analysis, aiding in precise prevention and treatment of cardiovascular disease in patients with diabetes.

Study design

The three key hypotheses of this MR analysis were as follows: (1) genetic variation in different subtypes of diabetes is strongly associated with cardiovascular disease; (2) genetic variation is independent of known confounders; and (3) genetic variation affects cardiovascular disease only through different subtypes of diabetes.

Data sources

This study analyzed pooled data from genome-wide association studies (GWAS), all of which are publicly accessible and approved by the appropriate ethical review boards. Data for the five diabetes subtypes were derived from GWAS and genetic risk score (GRS) analyses, involving 6986 individuals with various diabetes subtypes and 2744 controls [11]. Coronary heart disease data were derived from a meta-analysis of 60 801 cases and 123 504 controls [12]. Hypertension and atrial fibrillation data were obtained from the UK Biobank [13]. Heart failure data were derived from the HERMES consortium, including 47 309 cases and 930 014 controls [14]. Peripheral arterial disease data were sourced from the Japan Biobank (n=179 000) and GWAS from the UK Biobank and FinnGen (n=628 000) [15].

Selection and validation of single nucleotide polymorphisms

We screened for statistically significant single nucleotide polymorphisms (SNPs) as instrumental variables for subsequent studies using the following criteria.

(1) We defined the conditions for instrumental variables (IVs) as P < 1 × 10⁻⁵, a linkage disequilibrium threshold of R² < 0.001, and a clustering distance = 10 000 kb, to ensure that the IVs were independent of each other [16].

(2) The strength of correlation between the SNP loci and the exposure factors was estimated by calculating the f-statistic (β²/standard error²); when f-statistic was greater than 10, the effect of potential bias was negligible [17].

MR analysis

The main analytical method was inverse variance-weighted (IVW) analysis, which was used in conjunction with sensitivity analyses such as the weighted median and MR-Egger methods, to assess the robustness of the results. The weighted median method provided reliable causal estimates when at least 50% of the IVs in the analysis were valid [18]. The MR-Egger method assesses the horizontal polytropy of IVs and provides estimates that are corrected for polytropy [19]. Cochrane's Q-value assesses the degree of heterogeneity of the IVs [20]. Additionally, leave-one-out sensitivity analysis was performed to assess the possibility that individual SNPs may disproportionately affect the overall MR analysis. A Bonferroni correction threshold was employed to adjust for multiple examinations between the subtypes of diabetes and cardiovascular disease (five cardiovascular diseases), and the threshold for statistical significance was set at p < 0.01, with associations considered suggestive if the p-value was between 0.05 and 0.01. All statistical analyses were performed using the “TwoSampleMR” package in R version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria).

SNP selection and validation

This study was primarily based on the analyses of the GWAS database for European populations (Supplementary Table 1). All SNPs had an F-statistic of 10 or more, indicating that there were no weak instrumental variables in the analyses. The characteristics of the relevant SNPs are listed in Supplementary Table 2.

Mild obesity-related diabetes (MOD)

IVW analyses found (Figure 1) no positive or negative associations between MOD and the five cardiovascular diseases. Additionally, the weighted median and MR-Egger analyses (Supplementary Table 3) yielded consistent estimates. There was no evidence of direct pleiotropy.

Similar results were observed in the scatterplot (Supplementary Figure 1) and forest plot (Supplementary Figure 2) drawn to ascertain the correlation between MOD and cardiovascular disease. Omission sensitivity analyses (Supplementary Figure 3) revealed that individual SNPs did not have a disproportionate effect on the overall estimates. The funnel plot (Supplementary Figure 4) also showed no evidence of horizontal pleiotropy.

Mild age-related diabetes

IVW analyses found (Figure 2) no positive or negative associations between MARD and the five cardiovascular diseases. The weighted median and MR-Egger analyses (Supplementary Table 4) yielded consistent estimates. No evidence of directed pleiotropy was found.

Similar results were observed in the scatterplot (Supplementary Figure 5) and forest plot (Supplementary Figure 6) drawn to ascertain the association between MARD and cardiovascular disease. Omission sensitivity analyses (Supplementary Figure 7) revealed that individual SNPs did not have a disproportionate effect on the overall estimates. The funnel plot (Supplementary Figure 8) showed no evidence of horizontal pleiotropy.

Severe insulin-deficient diabetes

IVW analysis provided (Figure 3) suggestive evidence of a positive association between SIDD and peripheral arterial disease. Additionally, the weighted median and MR-Egger analyses yielded (Supplementary Table 5) consistent estimates. Notably, we found evidence of directed pleiotropy in patients with SIDD and peripheral arterial disease.

Similar results were observed in the scatter plot (Supplementary Figure 9) and forest plot (Supplementary Figure S10) for the association between SIDD and CVD. The omission sensitivity analysis showed (Supplementary Figure 11) that individual SNPs did not have an excessive effect on the overall estimate. Similar results were obtained from the funnel plot (Supplementary Figure 12).

Severe autoimmune diabetes

IVW analysis revealed (Figure 4) a positive association between SAID and peripheral arterial disease, with weighted median and MR-Egger analyses (Supplementary Table 6) showing consistent estimates. Furthermore, IVW analyses showed evidence of a negative association between SAID and atrial fibrillation. Additionally, targeted pleiotropy was observed in SAID with peripheral arterial disease.

Similar results were observed in the scatterplot (Supplementary Figure 13) and forest plot (Supplementary Figure 14) drawn to ascertain the association between SAID and CVD. Omission sensitivity analysis showed (Supplementary Figure 15) that individual SNPs did not have an excessive effect on the overall estimate. Similar results were obtained from the funnel plot (Supplementary Figure 16).

Severe insulin-resistant diabetes

IVW analyses showed (Figure 5) no significant correlation between SIRD and cardiovascular disease. The weighted median and MR-Egger analyses yielded (Supplementary Table 7) similar results.

The scatterplots, forest plots, omitted sensitivity analyses, and funnel plots for the association of SIRD with cardiovascular disease are presented in Supplementary Figures 17, 18, 19, and 20, respectively.

Recently, new subcomponent strata of diabetes have been proposed to provide individualized and precise treatments to patients with diabetes. However, the causal relationship between these diabetes subtypes and cardiovascular disease has not yet been effectively assessed. Given the high prevalence of cardiovascular disease in the population with diabetes, we conducted a two-sample MR study to assess the potential association between different subtypes of diabetes and cardiovascular disease. The principal findings were as follows: 1) SIDD has a potential positive causal association with PAD; and 2) SAID is positively and causally associated with peripheral arterial disease, and there is evidence suggestive of a negative causal association between SAID and atrial fibrillation.

The incidence of cardiovascular disease is significantly higher in patients with diabetes, and it has become one of the leading causes of death in these patients [21]. Diabetes is a major global public health concern. In 2018, researchers classified diabetes into five new subtypes using cluster analysis, laying the groundwork for precision medicine for diabetes [4]. This study aimed to determine the association between these five subtypes of diabetes and cardiovascular disease using two-sample MR analysis and further enhance our understanding of the relationship between diabetes and cardiovascular pathogenesis. Hernández et al. [22] conducted a cross-sectional study and included 71 patients with latent autoimmune diabetes in adults (LADA), 191 patients with type 2 diabetes mellitus, and 116 patients with type 1 diabetes mellitus. They found that carotid plaques were more frequent and multiple plaques were more common in patients with LADA than in the remaining two diabetic populations. This is consistent with the positive causal association between SAID and peripheral arterial observed in this study. The relationship between diabetes and the development of atrial fibrillation remains controversial [23]. In this study, we found that SAID might have a protective effect against the development of atrial fibrillation. Wei et al. [24] included 550 patients with autoimmune diabetes, 2,001 patients with type 2 diabetes, 1,573 patients with type 1 diabetes, and 2,355 individuals without diabetes in a case-control study and found that patients with autoimmune diabetes had an higher risk of cardiovascular morbidity; however, patients with diabetes with low autoimmune antibodies had a lower risk of cardiovascular morbidity than those with high autoimmune antibody diabetes. However, we did not observe a statistically significant association between SAID scores and coronary heart disease, hypertension, or heart failure, which may be explained by the variability in autoimmune levels between populations. Furthermore, a potentially positive causal association was observed between the SSID subtype and peripheral arterial disease.

Notably, in the analyses of SIRD, MOD, and MARD, we observed no significant correlation between the subtypes of diabetes and cardiovascular disease. Li et al. [5] conducted a retrospective cohort study that included 712 patients with diabetes mellitus and classified them into SAID, SIDD, SIRD, MOD, and MARD; they found a significant increase in cardiovascular mortality in patients with diabetes mellitus in each of these subgroups. This seems to contradict the results of the current study, and we hypothesized that this disparity could be attributed to the following: 1) a substantial amount of the data in this study came from a European population, and there may have been heterogeneity in the population; and 2) the sample size was insufficient, with an average of approximately 4 000 subtyped populations with diabetes included in the present study (Supplementary Table 1), which may have contributed to insufficient statistical power that contributed to these negative results.

Strengths and limitations

This study combined GWAS data to present more objective results for the causal relationship between subtypes of diabetes mellitus and cardiovascular disease and synthesized a variety of MR analyses, including IVW, weighted median, MR-Egger, and leave-one-out sensitivity analysis, and the findings were robust. Additionally, this study included multiple types of diabetes and cardiovascular disease, providing a more comprehensive insight into the cardiovascular–diabetes interrelationship.

The limitations of this study are as follows. We observed the presence of targeted pleiotropy; thus, the results may be affected by confounders, necessitating cautious interpretation. Furthermore, this study is based mainly on analyses of European populations; therefore, the generalizability of the results is limited. In the future, GWAS data should be analyzed on a larger scale and in multiple populations to exclude potential confounding factors and increase the reliability of the results.

There is a positive causal association between SAID and peripheral arterial disease and suggestive evidence of a negative association with atrial fibrillation. SSID has a potential positive causal association with peripheral arterial disease. This study facilitates further enhancement of the precise management of cardiovascular disease by clarifying interrelationships between the subtype of diabetes and cardiovascular disease.

Acknowledgments

We thank all the researchers who contributed to the study.

Conflict of Interest

The authors have no conflicts of interest.

Author Contributions

XYY: Writing–original draft. KL, CLY: Writing–review and editing. FZ, KMT: Data management, Methods. JJW, LC, HX, YXC: Formal analysis. YML, XWH, JFL: Validation, Visualization. YM: Supervision. All authors have read and approved the submitted version.

Funding

This work was supported by the Department of Science and Technology of Guizhou Province [grant number Qiankehe LH Zi [2017] 7109] and the Guizhou Maotai Group Hospital Scientific Research Programme Project (grant number [MTyk2024-01]).

Data Availability Statement

This study used publicly available datasets.

Roth GA, Forouzanfar MH, Moran AEet al. Demographic and epidemiologic drivers of global cardiovascular mortality. The New England journal of medicine 372: 1333-1341, 2015.
Ezzati M, Obermeyer Z, Tzoulaki I, Mayosi BM, Elliott P, Leon DA Contributions of risk factors and medical care to cardiovascular mortality trends. Nature reviews. Cardiology 12: 508-530, 2015.
Kovacic JC, Castellano JM, Farkouh ME, Fuster V The relationships between cardiovascular disease and diabetes: focus on pathogenesis. Endocrinology and metabolism clinics of North America 43: 41-57, 2014.
Ahlqvist E, Storm P, Käräjämäki Aet al. Novel subgroups of adult-onset diabetes and their association with outcomes: a data-driven cluster analysis of six variables. The lancet. Diabetes & endocrinology 6: 361-369, 2018.
Li PF, Chen WL Are the Different Diabetes Subgroups Correlated With All-Cause, Cancer-Related, and Cardiovascular-Related Mortality? The Journal of clinical endocrinology and metabolism 105, 2020.
Wang W, Jia T, Liu Yet al. Data-driven subgroups of newly diagnosed type 2 diabetes and the relationship with cardiovascular diseases at genetic and clinical levels in Chinese adults. Diabetes & metabolic syndrome 17: 102850, 2023.
Davey Smith G, Ebrahim S What can mendelian randomisation tell us about modifiable behavioural and environmental exposures? BMJ (Clinical research ed.) 330: 1076-1079, 2005.
Higbee DH, Granell R, Sanderson E, Davey Smith G, Dodd JW Lung function and cardiovascular disease: a two-sample Mendelian randomisation study. The European respiratory journal 58, 2021.
Rolver MG, Emanuelsson F, Nordestgaard BG, Benn M Contributions of elevated CRP, hyperglycaemia, and type 2 diabetes to cardiovascular risk in the general population: observational and Mendelian randomization studies. Cardiovascular diabetology 23: 165, 2024; Liu Z, Wang H, Yang Z, Lu Y, Zou C Causal associations between type 1 diabetes mellitus and cardiovascular diseases: a Mendelian randomization study. Cardiovascular diabetology 22: 236, 2023.
Ruan Z, Zhao J Differential ischemic stroke risk linked to novel subtypes of type 2 diabetes: insights from a Mendelian randomization analysis. Endocrine, 2024.
Mansour Aly D, Dwivedi OP, Prasad RBet al. Genome-wide association analyses highlight etiological differences underlying newly defined subtypes of diabetes. Nature genetics 53: 1534-1542, 2021.
Nikpay M, Goel A, Won HHet al. A comprehensive 1,000 Genomes-based genome-wide association meta-analysis of coronary artery disease. Nature genetics 47: 1121-1130, 2015.
Dönertaş HM, Fabian DK, Valenzuela MF, Partridge L, Thornton JM Common genetic associations between age-related diseases. Nature aging 1: 400-412, 2021; . Nielsen JB, Thorolfsdottir RB, Fritsche LGet al. Biobank-driven genomic discovery yields new insight into atrial fibrillation biology. Nature genetics 50: 1234-1239, 2018.
Shah S, Henry A, Roselli Cet al. Genome-wide association and Mendelian randomisation analysis provide insights into the pathogenesis of heart failure. Nature communications 11: 163, 2020.
Sakaue S, Kanai M, Tanigawa Yet al. A cross-population atlas of genetic associations for 220 human phenotypes. Nature genetics 53: 1415-1424, 2021.
Ning J, Huang SY, Chen SD, Zhang YR, Huang YY, Yu JT Investigating Casual Associations Among Gut Microbiota, Metabolites, and Neurodegenerative Diseases: A Mendelian Randomization Study. Journal of Alzheimer's disease : JAD 87: 211-222, 2022.
Triozzi JL, Hsi RS, Wang Get al. Mendelian Randomization Analysis of Genetic Proxies of Thiazide Diuretics and the Reduction of Kidney Stone Risk. JAMA network open 6: e2343290, 2023.
Bowden J, Davey Smith G, Haycock PC, Burgess S Consistent Estimation in Mendelian Randomization with Some Invalid Instruments Using a Weighted Median Estimator. Genetic epidemiology 40: 304-314, 2016.
Bowden J, Davey Smith G, Burgess S Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. International journal of epidemiology 44: 512-525, 2015.
Greco MF, Minelli C, Sheehan NA, Thompson JR Detecting pleiotropy in Mendelian randomisation studies with summary data and a continuous outcome. Statistics in medicine 34: 2926-2940, 2015.
Dal Canto E, Ceriello A, Rydén Let al. Diabetes as a cardiovascular risk factor: An overview of global trends of macro and micro vascular complications. European journal of preventive cardiology 26: 25-32, 2019.
Hernández M, López C, Real Jet al. Preclinical carotid atherosclerosis in patients with latent autoimmune diabetes in adults (LADA), type 2 diabetes and classical type 1 diabetes. Cardiovascular diabetology 16: 94, 2017.
Johansson C, Örtendahl L, Lind MM, Andersson J, Johansson L, Brunström M Diabetes, prediabetes, and atrial fibrillation-A population-based cohort study based on national and regional registers. Journal of internal medicine 294: 605-615, 2023.
Wei Y, Herzog K, Ahlqvist Eet al. All-Cause Mortality and Cardiovascular and Microvascular Diseases in Latent Autoimmune Diabetes in Adults. Diabetes care 46: 1857-1865, 2023.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Association of new subtypes of diabetes with cardiovascular disease: Insights from Mendelian randomization

Status:

Version 1

Abstract

Figures

Introduction

Materials and Methods

Study design

Data sources

Selection and validation of single nucleotide polymorphisms

MR analysis

Results

SNP selection and validation

Mild obesity-related diabetes (MOD)

Mild age-related diabetes

Severe insulin-deficient diabetes

Severe autoimmune diabetes

Severe insulin-resistant diabetes

Discussion

Strengths and limitations

Conclusions

Declarations

Acknowledgments

Conflict of Interest

Author Contributions

Funding

Data Availability Statement

References

Additional Declarations

Supplementary Files

Status:

Version 1