OBJECTIVE

To investigate the causal association of type 2 diabetes and its components with risk of vascular complications independent of shared risk factors obesity and hypertension and to identify the main driver of this risk.

RESEARCH DESIGN AND METHODS

We conducted Mendelian randomization (MR) using independent genetic variants previously associated with type 2 diabetes, fasting glucose, HbA1c, fasting insulin, BMI, and systolic blood pressure as instrumental variables. We obtained summary-level data for 18 vascular diseases (15 for type 2 diabetes) from FinnGen and publicly available genome-wide association studies as our outcomes. We conducted univariable and multivariable MR, in addition to sensitivity tests to detect and minimize pleiotropic effects.

RESULTS

Univariable MR analysis showed that type 2 diabetes was associated with 9 of 15 outcomes; BMI and systolic blood pressure were associated with 13 and 15 of 18 vascular outcomes, respectively; and fasting insulin was associated with 4 and fasting glucose with 2. No robust association was found for HbA1c instruments. With adjustment for correlated traits in the multivariable test, BMI and systolic blood pressure, consistent causal effects were maintained, while five associations with type 2 diabetes (chronic kidney disease, ischemic heart disease, heart failure, subarachnoid hemorrhage, and intracerebral hemorrhage) were attenuated to null.

CONCLUSIONS

Our findings add strong evidence to support the importance of BMI and systolic blood pressure in the development of vascular complications in people with type 2 diabetes. Such findings strongly support the need for better weight and blood pressure management in type 2 diabetes, independent of glucose lowering, to limit important complications.

The link between type 2 diabetes and vascular complications has been subject to extensive research in both observational studies and randomized control trials with somewhat mixed findings. Multiple guidelines suggest tight glycemic control is critical to protection against macrovascular and microvascular complications associated with diabetes (1). However, findings from randomized control trials show less significant benefits from tight glycemic control than many expected, at least in the short-term to medium term (2).

The complex and heterogenic nature of type 2 diabetes makes it hard to reach a clear-cut conclusion regarding its role in development of vascular complications. For instance, type 2 diabetes affects a broad range of organs within the human body, and despite the mounting evidence provided by observational studies indicating its role in development of several macrovascular and microvascular complications, to some extent, the effect of several tightly correlated risk factors, such as obesity and hypertension, makes the role of type 2 diabetes and glycemic traits in development of those complications uncertain (3). Another source of complexity is the heterogeneity in underlying mechanisms. Type 2 diabetes may occur due to resistance to insulin actions in the insulin-sensitive tissues such as the liver, muscles, and adipose tissues combined with insufficient insulin secretion as a result of β-cells dysfunction, both occurring in the early stages of the disease (4). Hence, with observational studies one cannot determine relative contributions of differing risk pathways to diabetes complications. Also, excess adiposity and higher blood pressure (5) are known to predate the diagnosis of diabetes, sometimes by many years, with excess adiposity an upstream pathogenic factor for diabetes development in many (6). Plus, there is the issue of aggregated weight exposure, which could drive many vascular complications (7), a concept easily overlooked even though most may have many years of excess adiposity before diagnosis of type 2 diabetes.

In this study, we used Mendelian randomization (MR), a method that leverages genetic variation to establish causal relationships between modifiable risk factors and outcomes (8), to investigate causal association between type 2 diabetes and a range of macrovascular and microvascular complications. To elucidate the specific contribution of type 2 diabetes and shared risk factors to these associations, we conducted multivariable MR (MVMR), adjusting for the genetically predicted effect of BMI and systolic blood pressure. We additionally looked at causal effects of higher HbA1c, fasting insulin, and fasting glucose in the nondiabetes range against the risk of same vascular complications. A better understanding of the causal risk factor driving vascular complications in individuals with type 2 diabetes and those with prediabetes could inform the development of targeted prevention strategies.

Study Design

We designed an MR study to dissect the causal role of type 2 diabetes in vascular complications (Supplementary Fig. 1). MR is a statistical method where genetic variants are used as instrumental variables to estimate the causal effect of an exposure (e.g., type 2 diabetes) on an outcome (e.g., vascular disease). Since genetic variants are randomly assorted at conception, this method can significantly reduce confounding and reverse causation.

We used genetic instruments for type 2 diabetes, three glycemic traits in the nondiabetes range (fasting glucose, HbA1c, and fasting insulin), BMI, and systolic blood pressure as our exposures and 18 vascular complications as outcomes in a univariable MR model (Supplementary Fig. 1A). We excluded diabetic nephropathy, neuropathy, and retinopathy from the analysis of type 2 diabetes and in MVMR models to avoid collider bias, as these complications follow the diagnosis of type 2 diabetes. Since obesity and hypertension are shared risk factors and can confound the association between type 2 diabetes and vascular complications, we performed MVMR to try to exclude the confounding effect of BMI and systolic blood pressure (Supplementary Fig. 1B). The causal effects as estimated with MVMR differ from those estimated with the univariable method. Univariable MR estimates the total causal effect of exposure on an outcome, while the multivariable method estimates the independent direct causal effect of each exposure on the outcome of interest (9).

Genetic Instrument Selection

To construct a genetic instrument for type 2 diabetes, we used a recent genome-wide association study (GWAS) of type 2 diabetes (10) (74,124 case and 824,006 control subjects). For glycemic traits fasting glucose, fasting insulin, and HbA1c, we obtained the genetic instruments from the Meta-Analyses of Glucose and Insulin-related traits Consortium (MAGIC) (281,416 individuals without diabetes) (11). The association tests for fasting glucose and fasting insulin in the GWAS included adjustment for BMI, which may produce collider bias. Therefore, to ensure that the issue of collider bias was addressed, we performed a sensitivity analysis using instruments obtained from BMI-unadjusted GWAS (140,595 and 98,210 individuals without diabetes for fasting glucose and fasting insulin, respectively) (12). For BMI instrument we used data from a GWAS of 694,649 individuals (13) and for systolic blood pressure from the GWAS of 757,601 individuals (14). All GWAS were in individuals of European ancestry.

To create the genetic instruments, we first selected variants associated with each exposure (P < 5 × 10−8) and then identified a set of independent variants for each exposure using linkage disequilibrium pruning (r2 > 0.001) within a window of 10 Mb with inclusion of unrelated white Europeans from the 1000 Genomes reference panel. A summary of each exposure GWAS, including definitions, numbers of case and control subjects, and covariates adjusted for in the GWAS model, can be found in Supplementary Table 1.

For MVMR, we used genetic variants from the same GWAS of exposure data and divided them into four instruments. Each instrument included genetic variants independently associated with BMI and systolic blood pressure in addition to one of type 2 diabetes, fasting glucose, HbA1c, and fasting insulin. The purpose of this step was to understand the causal effect of each exposure after adjustment for the genetically predicted higher BMI and systolic blood pressure.

Outcome Data Sources

We selected a broad range of diabetes-associated cardiovascular (including stroke, heart failure, and atrial fibrillation) and microvascular (including chronic kidney disease, diabetic retinopathy, diabetic nephropathy, diabetic neuropathy, and retinal vascular occlusion) outcomes. We obtained genome-wide summary-level data for 18 outcomes from FinnGen consortium release 6 (https://www.finngen.fi/en/). For six outcomes (ischemic stroke [15], ischemic heart disease [16], heart failure [17], atrial fibrillation [18], myocardial infarction [19], and chronic kidney disease [20]), data from another independent published GWAS were available. For these six outcomes we meta-analyzed results from FinnGen and published GWAS. We obtained all outcomes with use of the GWAS database (available from https://gwas.mrcieu.ac.uk/) of the MRC Integrative Epidemiology Unit at the University of Bristol to ensure the homogeneity of study population and accuracy of the obtained result. Detailed information on outcome data sources (e.g., outcome definitions, numbers of case and control subjects, and covariates adjusted for in the GWAS model) can be found in Supplementary Table 2.

Statistical Analysis

We used different MR methods. A robust MR study must satisfy three assumptions: first, the relevance assumption, where the instrument variables must be associated with the exposure; second, the independence assumption, which states that instrument variable must have no association with a confounder; third, the exclusion restriction assumption, which suggests that the relation between the instrument variable and the outcome should only be through the exposure. Invalidation of any of those three core assumptions could invalidate the MR. Therefore, the three assumptions must be evaluated in advance (21).

For the main analysis, we used the inverse variance–weighted (IVW) method. The IVW method combines the Wald ratio estimates of each single nucleotide polymorphism into one causal estimate. However, the IVW result could be affected by instrumental variable bias or horizontal pleiotropy. An instrumental variable with weaker association with the exposure tends to produce bias in the direction of the observational confounded association proportional to the strength of the association. On the other hand, horizontal pleiotropy occurs when an instrument variable has direct effect on the outcome that is bypassing the exposure via a different pathway to the exposure, violating MR’s third assumption. Therefore, we used MR-Egger as a method of sensitivity testing and for determining presence of horizontal pleiotropy based on Egger intercept in addition to weighted median, simple mode, and weighted mode.

For the estimation of the independent effect of each exposure, we used MVMR IVW method. This method uses several phenotypes as one exposure into the model. Since type 2 diabetes, BMI, and systolic blood pressure are correlated, we created a model that includes all of these three exposures to identify which one drives the risk of vascular complications.

All MR analyses were performed with R (version 4.2.2). The univariable MR and MVMR were conducted with the TwoSampleMR package (22). We used the metafor package for meta-analysis of MR results from FinnGen and published GWAS.

We used Benjamini-Hochberg–adjusted P values (BHP) to classify significant causal associations (BHP <0.05). Those associations with IVW P value <0.05 and BHP >0.05 are presented as suggestive associations.

Data and Resource Availability

All data used in this paper are accessible through IEU OpenGWAS project database, available from https://gwas.mrcieu.ac.uk/, and FinnGen R6, available from https://www.finngen.fi/en/access_results. All data generated in this study are included in the published article and its online supplementary files.

To conduct univariable MR, we selected 186, 70, 75, 38, 456, and 543 genetic variants associated with type 2 diabetes, fasting glucose, HbA1c, fasting insulin, systolic blood pressure and BMI, respectively, as instrumental variables. The F statistics for these instrument were between 51 and 122, indicating the instrument strength (23) (Supplementary Table 3).

Causal Effect of Type 2 Diabetes on Risk of Vascular Complications

Genetically predicted type 2 diabetes was associated with 9 of the 15 cardiovascular and microvascular outcomes including ischemic stroke, ischemic heart disease, heart failure, myocardial infarction, chronic kidney disease, peripheral atherosclerosis, and peripheral artery disease (Fig. 1 and Supplementary Table 4). There was also a suggestive protective effect against aortic aneurysm (odds ratio [OR] 0.91 [95% CI 0.85–0.99]; P = 0.024) (Fig. 1, Supplementary Fig. 2, and Supplementary Table 4). MR-Egger sensitivity analysis indicated no evidence of horizontal pleiotropy (Supplementary Table 5). The effects from all the sensitivity tests were consistent with the IVW estimates (Supplementary Table 6). After adjustment for the genetically predicted effect of BMI and systolic blood pressure in the MVMR test, genetically predicted type 2 diabetes lost its effect on chronic kidney disease, ischemic heart disease, heart failure, subarachnoid hemorrhage, and intracerebral hemorrhage. However, an association with a diluted effect size remained with ischemic stroke and myocardial infarction, while the association and the effect size with peripheral atherosclerosis and peripheral artery disease remained unchanged (Fig. 1 and Supplementary Table 7).

Figure 1

Comparison of the causal effect of type 2 diabetes, BMI, and systolic blood pressure from univariable MR (UVMR) and MVMR tests: the plot shows the total causal effect of type 2 diabetes, BMI, and systolic blood pressure on 15 and 18 vascular complications, respectively, and the independent causal effect of each exposure after adjustment for the genetically predicted effect of the other two. The color and intensity of the color correspond with the direction and value of z scores (from IVW test). BHP <0.05 are given. *Meta-analyzed results from FinnGen and a published GWAS were available. NA, not applicable; SBP, systolic blood pressure; T2D, type 2 diabetes.

Figure 1

Comparison of the causal effect of type 2 diabetes, BMI, and systolic blood pressure from univariable MR (UVMR) and MVMR tests: the plot shows the total causal effect of type 2 diabetes, BMI, and systolic blood pressure on 15 and 18 vascular complications, respectively, and the independent causal effect of each exposure after adjustment for the genetically predicted effect of the other two. The color and intensity of the color correspond with the direction and value of z scores (from IVW test). BHP <0.05 are given. *Meta-analyzed results from FinnGen and a published GWAS were available. NA, not applicable; SBP, systolic blood pressure; T2D, type 2 diabetes.

Close modal

Genetically predicted higher BMI was associated with higher risk of 13 out of 18 cardiovascular and microvascular diseases, including ischemic stroke, ischemic heart disease, diabetic retinopathy, diabetic nephropathy, and diabetic neuropathy. Outcomes for which we found no evidence of a causal association included retinal hemorrhage, retinal vascular occlusion, subarachnoid hemorrhage, cerebral aneurysm, and nonruptured and intracerebral hemorrhage (Fig. 1, Supplementary Fig. 3, and Supplementary Table 8). These effects were consistent across all sensitivity tests (Supplementary Table 9), and we found no evidence of horizontal pleiotropy from the MR-Egger intercept (Supplementary Table 10). In the multivariable model, adjustment for the genetically predicted effect of type 2 diabetes and systolic blood pressure removed association with peripheral artery disease (Fig. 1 and Supplementary Table 7) but not other outcomes.

Genetically predicted higher systolic blood pressure was associated with higher risk of 15 of the 18 vascular outcomes, including ischemic stroke, cerebral aneurysm, ischemic heart disease, and myocardial infarction (Fig. 1, Supplementary Fig. 4, and Supplementary Table 11). These effects were consistent across all sensitivity test (Supplementary Table 12), and we found no evidence of horizontal pleiotropy from the MR-Egger intercept (Supplementary Table 13). Adjustment for the genetically predicted effect of BMI and type 2 diabetes removed the association with pulmonary embolism (Fig. 1 and Supplementary Table 7) and attenuated the effect for chronic kidney disease, heart failure, ischemic heart disease, subarachnoid hemorrhage, and intracerebral hemorrhage toward the null.

Causal Effect of Higher Levels of Glycemic Traits in Their Normal Range on Risk of Vascular Complications

A 1 mmol/L increase of genetically predicted fasting glucose was associated with higher risk of ischemic heart disease (OR 1.26 [95% CI 1.12–1.41]; BHP = 2 × 10−3) and higher risk of diabetic nephropathy (OR 1.98 [1.35–2.91]; BHP = 8 × 10−3). There was suggestive evidence of association with diabetic retinopathy (1.36 [1.06–1.75]), peripheral atherosclerosis (1.68 [1.18–2.38]), and peripheral artery disease (1.57 [1.15–2.16]) (Fig. 2, Supplementary Fig. 5, and Supplementary Table 14). These effects were consistent across all sensitivity tests (Supplementary Table 15), and we found no evidence of horizontal pleiotropy from the MR-Egger intercept (Supplementary Table 16). In the multivariable model, with adjustment for the genetically predicted effect of BMI and systolic blood pressure, the effect on ischemic heart disease was attenuated toward the null, while the suggestive association with peripheral atherosclerosis and peripheral artery disease remained unchanged (Fig. 2). Adjustment for the genetically predicted effect of fasting glucose in the multivariable model did not change the associations between BMI or systolic blood pressure and vascular outcomes.

Figure 2

Comparison of the causal effect of fasting glucose, BMI, and systolic blood pressure from univariable MR (UVMR) and MVMR tests. The plot shows the total causal effect of fasting glucose, BMI, and systolic blood pressure on 18 vascular complications and the independent causal effect of each exposure after adjustment for the genetically predicted effect of the other two. The color and intensity of the color correspond with the direction and value of z scores (from IVW test). BHP <0.05 are given. *Meta-analyzed results from FinnGen and a published GWAS were available. FG, fasting glucose; SBP, systolic blood pressure.

Figure 2

Comparison of the causal effect of fasting glucose, BMI, and systolic blood pressure from univariable MR (UVMR) and MVMR tests. The plot shows the total causal effect of fasting glucose, BMI, and systolic blood pressure on 18 vascular complications and the independent causal effect of each exposure after adjustment for the genetically predicted effect of the other two. The color and intensity of the color correspond with the direction and value of z scores (from IVW test). BHP <0.05 are given. *Meta-analyzed results from FinnGen and a published GWAS were available. FG, fasting glucose; SBP, systolic blood pressure.

Close modal

There was no multiple testing–adjusted evidence for an association with HbA1c, but we observed suggestive association between 1% increase in genetically predicted HbA1c and higher risk of ischemic heart disease (OR 1.26 [95% CI 1.04–1.51]), diabetic retinopathy (1.54 [1.14–2.07]), and diabetic nephropathy (1.87 [1.16–3.02]) (Fig. 3, Supplementary Fig. 6, and Supplementary Table 17). These effects were consistent across all sensitivity tests (Supplementary Table 18), and we found no evidence of horizontal pleiotropy from the MR-Egger intercept (Supplementary Table 19). After adjustment for the genetically predicted effect of BMI and systolic blood pressure in the multivariable model, genetically predicted HbA1c suggestive association with ischemic heart disease remained plus a new suggestive association with myocardial infarction (1.26 [1.05–1.67]). Adjustment for the genetically predicted effect of HbA1c in the multivariable model did not change the associations between BMI or systolic blood pressure and vascular outcomes.

Figure 3

Comparison of the causal effect of HbA1c, BMI, and systolic blood pressure from univariable MR (UVMR) and MVMR tests. The plot shows the total causal effect of HbA1c, BMI, and systolic blood pressure on 18 vascular complications and the independent causal effect of each exposure after adjustment for the genetically predicted effect of the other two. The color and intensity of the color correspond with the direction and value of z scores (from IVW test). BHP <0.05 are given. *Meta-analyzed results from FinnGen and a published GWAS were available. SBP, systolic blood pressure.

Figure 3

Comparison of the causal effect of HbA1c, BMI, and systolic blood pressure from univariable MR (UVMR) and MVMR tests. The plot shows the total causal effect of HbA1c, BMI, and systolic blood pressure on 18 vascular complications and the independent causal effect of each exposure after adjustment for the genetically predicted effect of the other two. The color and intensity of the color correspond with the direction and value of z scores (from IVW test). BHP <0.05 are given. *Meta-analyzed results from FinnGen and a published GWAS were available. SBP, systolic blood pressure.

Close modal

A 1 pmol/L increase in genetically predicted fasting insulin was associated with higher risk of ischemic heart disease (OR 1.88 [95% CI 1.45–2.44]), myocardial infarction (2.06 [1.55–2.73]), diabetic nephropathy (3.86 [1.87–7.98]), and diabetic neuropathy (4.76 [1.85–12.26]) (Fig. 4, Supplementary Fig. 7, and Supplementary Table 20). There was also evidence of suggestive association with higher risk of ischemic stroke (1.35 [1.05–1.75]), chronic kidney disease (1.63 [1.15–2.31]), peripheral atherosclerosis (2.60 [1.32–5.13]), and peripheral artery disease (2.41 [1.31–4.44]) (Fig. 4 and Supplementary Fig. 7). These effects were consistent across all sensitivity tests (Supplementary Table 21), and we found no evidence of horizontal pleiotropy from the MR-Egger intercept test (Supplementary Table 22). With adjustment for the genetically predicted effect of BMI and SBP in the multivariable model, genetically predicted fasting insulin lost its association with ischemic heart disease, while its association with myocardial infarction was downgraded toward suggestive, as the effect size attenuated toward the null (Fig. 4). Adjustment for the genetically predicted effect of fasting insulin in the multivariable model did not change the associations between BMI or systolic blood pressure and vascular outcomes.

Figure 4

Comparison of the causal effect of fasting insulin, BMI, and systolic blood pressure from univariable MR (UVMR) and MVMR tests. The plot shows the total causal effect of fasting insulin, BMI, and systolic blood pressure on 18 vascular complications and the independent causal effect of each exposure after adjustment for the genetically predicted effect of the other two. The color and intensity of the color correspond with the direction and value of z scores (from IVW test). BHP <0.05 are given. *Meta-analyzed results from FinnGen and a published GWAS were available. FI, fasting insulin; SBP, systolic blood pressure.

Figure 4

Comparison of the causal effect of fasting insulin, BMI, and systolic blood pressure from univariable MR (UVMR) and MVMR tests. The plot shows the total causal effect of fasting insulin, BMI, and systolic blood pressure on 18 vascular complications and the independent causal effect of each exposure after adjustment for the genetically predicted effect of the other two. The color and intensity of the color correspond with the direction and value of z scores (from IVW test). BHP <0.05 are given. *Meta-analyzed results from FinnGen and a published GWAS were available. FI, fasting insulin; SBP, systolic blood pressure.

Close modal

We conducted an MR study to try to investigate the causal association of type 2 diabetes, BMI, and systolic blood pressure with cardiovascular and microvascular complications. Type 2 diabetes, BMI, and systolic blood pressure genetic instruments were associated with most of the outcomes tested in a univariable analysis. However, in a multivariable model the associations between genetically predicted type 2 diabetes and several of the vascular outcomes attenuated to the null, while the majority of associations between genetically determined BMI and systolic blood pressure remained significant. To understand the causal role of “glycemic” traits in their nondiabetes range, we followed the same approach. Genetically predicted fasting insulin showed the most significant causal association with vascular outcomes in the univariable model, but notably all causal effects disappeared with correction for the effect of genetically predicted higher BMI and systolic blood pressure.

Type 2 Diabetes and Vascular Complications

Our findings are consistent with previous observational (24,25) and MR studies (3,26) highlighting type 2 diabetes as a causal risk factor for most of the common cardiovascular and microvascular complications including peripheral artery disease, peripheral atherosclerosis, myocardial infarction, heart failure, chronic kidney disease, ischemic stroke, and ischemic heart disease. We additionally provided evidence that the causal role of type 2 diabetes in mechanisms that lead to these complications is independent of higher BMI and systolic blood pressure. The association with chronic kidney disease, heart failure, ischemic heart disease, subarachnoid hemorrhage, and intracerebral hemorrhage attenuated to the null after adjustment for the genetically predicted effect of BMI and systolic blood pressure, suggesting that the association reported in the observational studies for the link between type 2 diabetes and these complications may largely be mediated through obesity and hypertension.

Our results indicated a suggestive protective effect of type 2 diabetes against aortic aneurysm, consistent with observations from animal models and human studies suggesting that diabetes exerts protective effect against aortic aneurysms (27). There was no evidence of such effect against cerebral aneurysm. We also found no evidence of a causal association between genetically predicted type 2 diabetes and atrial fibrillation, consistent with findings reported from previous MR studies (3,28). By contrast, BMI instruments were strongly linked to this outcome.

Glycemic Traits and Vascular Complications

Hyperglycemia has long been linked to the progression of diabetes-associated microvascular complications, and investigators of several studies recommended intensive glycemic control as a protective approach (29). We did not find any robust evidence for a causal role of hyperglycemia in the normal range and higher risk of vascular complications. It is known that risk for many complications accelerates meaningfully only once glucose or HbA1c levels move into the diabetes range (30). We did not find any association between genetically predicted fasting glucose and the risk of chronic kidney disease, consistent with findings of previous studies. These findings suggest that the impact of glucose on chronic kidney disease may be a threshold effect with impact only once evidence of frank diabetes develops.

Our findings are consistent with a previous report on a causal link between fasting insulin and increased risk of coronary artery disease, ischemic stroke, and myocardial infarction (31). Observational studies have generally not focused on the association of fasting insulin and insulin resistance with cardiovascular and microvascular complications because of the general assumption that hyperglycemia is the main reason for the development of those complications. However, our genetic findings suggest that the positive association of fasting insulin as an independent risk factor is greater than that of hyperglycemic markers. Of course, this does not mean insulin per se is harmful—the results of the Outcome Reduction With Initial Glargine Intervention (ORIGIN Trial) (32) did not suggest harm for basal insulin in diabetes—but the factors that lead to higher insulin levels (i.e., tissue insulin resistance due to ectopic fat) may be harmful or else related factors such as dyslipidemia or lower activity levels may be relevant.

BMI, Systolic Blood Pressure, and Vascular Complications

In the univariable MR, genetically predicted higher BMI was causally associated with the majority of the vascular complications including diabetic nephropathy, diabetic retinopathy, and diabetic neuropathy.

In the MVMR, we excluded retinopathy, nephropathy, and neuropathy to avoid bias. The associations remained significant for 8 of 13 outcomes after adjustment for the genetically predicted effect of type 2 diabetes and systolic blood pressure, suggesting an independent causal role for higher BMI in diabetes-related vascular complications, but the association was lost for peripheral artery disease and peripheral atherosclerosis. These results are consistent with those reported previously (33), but in contrast, we did not find an independent causal effect of BMI on subarachnoid hemorrhage. These results are also consistent with the fact that many individuals will have had excess adiposity for a long period of time before frank type 2 diabetes develops, such that overall exposure to excess adiposity would have been extensive, operating over many years to accelerate the risk of many adiposity-sensitive complications, before and after diabetes develops (7). This is an important fact given that chronic complications like chronic kidney disease take time to develop from aggregated exposure to risk factors such as excess adiposity (34). Interestingly, in a recently reported 10-year observational follow-up of a trial comparing two forms of bariatric surgery with medical treatment in diabetes, the difference in weights by ∼20 kg in surgery recipients versus medical treated participants for ∼10 years was associated with a remarkable apparent difference in macrovascular/microvascular complication rates (6% vs. 71%, respectively) (35). Of course, this was a small study and the results of the ongoing SURPASS-CVOT in >12,000 patients with type 2 diabetes will be interesting, given it is testing the cardiovascular benefits of one incretin-based therapy (tirzepatide) that yields an ∼10 kg greater weight loss than another incretin-based therapy (dulaglutide) (36). This trial should report in 2 years’ time.

Genetically predicted higher systolic blood pressure was associated with a majority of the outcomes. All of these associations remained significant when we corrected for the genetically instrumented effect of type 2 diabetes and BMI. Our findings support those reported from various previous observational studies that suggested that cardiovascular complications are common among people with type 2 diabetes and hypertension, while microvascular complications risk is induced by hypertension (37). An independent study investigated the causal effect of hypertension on risk of cardiovascular diseases, and investigators found that 10 mmHg increase in genetically predicted systolic blood pressure increased risk of total cardiovascular disease (OR 1.32 [95% CI 1.25–1.40]), ischemic heart disease (OR 1.33 [1.24–1.41]), and stroke (OR 1.35 [1.24–1.48]) (38), which is in line with the findings of our study. These findings also fit with evidence for the benefit of blood pressure reduction in diabetes for many vascular complications (39).

Evidence From Non-European Populations

Limited availability of genome-wide association data has resulted in a scarcity of MR studies conducted on non-European populations. Nevertheless, in a notable MR study investigators examined 45 risk factors for chronic kidney disease in both European and East Asian populations (40). The results revealed that genetically predicted type 2 diabetes was associated with increased risk of chronic kidney disease in European, Japanese, and Chinese populations, while genetically predicted BMI was found to be associated with increased risk of chronic kidney disease in European and Japanese populations but not in the Chinese population (40). This inconsistency may be attributed to a lack of chronic kidney disease cases among the Chinese population or potential ethnic variations (41). Furthermore, genetically predicted systolic blood pressure was associated with chronic kidney disease in Europeans but not in the East Asian population, suggesting a potential effect based on ancestry (41). Another study investigated the causal effect of type 2 diabetes on risk of coronary artery disease and atrial fibrillation in the East Asian population and revealed a causal association with coronary artery disease but not with atrial fibrillation (42).

Strengths and Limitations

This study covered a broad range of vascular complications associated with type 2 diabetes. We used strong instrumental variables associated with each exposure (F statistic >10), which indicates good strength of our genetic instruments. The novelty of our study arises from the fact that we investigated the causal effect of type 2 diabetes and glycemic traits while looking at the mediating effect of the shared risk factors: BMI and systolic blood pressure appear to be lacking. We included genetic instruments for type 2 diabetes, BMI, and systolic blood pressure in a multivariable model to adjust for the genetically predicted independent effect of each phenotype and try to help answer the question of which component drives the risk of vascular complications. Our study had some limitations. First, our source of data was restricted to individuals of European ancestry, which makes the generalizability of our findings to other ethnic groups unclear. Given the excess risk of type 2 diabetes and vascular complications in non-Europeans, we hope the availability of non-European GWAS will make it possible to follow up our findings. Second, the pleiotropic effect of genetic variants we used as instrument could violate the MR assumption. To address this, we performed various sensitivity analyses and adjusted for the correlation of type 2 diabetes, BMI, and systolic blood pressure in our MVMR test. Third, we had data from two different sources, FinnGen and an independent published GWAS, for only 6 of 19 vascular outcomes where the same ICD definition (ICD-10) was used to define case and control subjects. For the other 13 outcomes where results were only available in FinnGen, it would be necessary to validate the findings in non-Finnish populations. Finally, our work on glycemic trait instruments was necessarily restricted to the nondiabetes range, and elevated glucose beyond the diabetes diagnostic thresholds accelerates vascular and kidney harm over many years, as is seen in patients living with type 1 diabetes but without obesity. This means that these data are somewhat limited.

Conclusion

We provided genetic evidence for causal effects of type 2 diabetes, BMI, and systolic blood pressure on risk of vascular complications. Our findings provide evidence that even though tight glycemic control is considered important for lowering the risk of vascular complications, such an approach alone is unlikely to be enough. Rather, additional weight and blood pressure management should have meaningful impacts to lower the risk of multiple complications, including heart failure, important arrhythmias, and chronic kidney disease, in those living with type 2 diabetes.

This article contains supplementary material online at https://doi.org/10.2337/figshare.23638977.

Acknowledgments. The authors thank the participants and investigators of the FinnGen study.

Funding. H.Y. is funded by Diabetes UK RD Lawrence fellowship (grant 17/0005594). N.S. is supported by the British Heart Foundation Research Excellence Award (RE/18/6/34217).

Duality of Interest. N.S. has received grants and personal fees from AstraZeneca, Boehringer Ingelheim, and Novartis; a grant from Roche Diagnostics; and personal fees from Abbott Laboratories, Afimmune, Amgen, Eli Lilly, Hanmi Pharmaceutical, Merck Sharp & Dohme, Novo Nordisk, Pfizer, and Sanofi outside the submitted work. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. A.A. analyzed data. H.Y. designed the study. A.A., N.S., and H.Y. wrote the manuscript. H.A. and F.D. reviewed and edited the manuscript. H.Y. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented in abstract form at Diabetes UK Professional Conference 2023, Liverpool, U.K., 26–28 April 2023.

1.
Zoungas
S
,
Arima
H
,
Gerstein
HC
, et al.;
Collaborators on Trials of Lowering Glucose (CONTROL) group
.
Effects of intensive glucose control on microvascular outcomes in patients with type 2 diabetes: a meta-analysis of individual participant data from randomised controlled trials
.
Lancet Diabetes Endocrinol
2017
;
5
:
431
437
2.
Ray
KK
,
Seshasai
SRK
,
Wijesuriya
S
, et al
.
Effect of intensive control of glucose on cardiovascular outcomes and death in patients with diabetes mellitus: a meta-analysis of randomised controlled trials
.
Lancet
2009
;
373
:
1765
1772
3.
Huang
M
,
Laina-Nicaise
LD
,
Zha
L
,
Tang
T
,
Cheng
X
.
Causal association of type 2 diabetes mellitus and glycemic traits with cardiovascular diseases and lipid traits: a Mendelian randomization study
.
Front Endocrinol (Lausanne)
2022
;
13
:
840579
4.
Zheng
Y
,
Ley
SH
,
Hu
FB
.
Global aetiology and epidemiology of type 2 diabetes mellitus and its complications
.
Nat Rev Endocrinol
2018
;
14
:
88
98
5.
Emdin
CA
,
Anderson
SG
,
Woodward
M
,
Rahimi
K
.
Usual blood pressure and risk of new-onset diabetes: evidence from 4.1 million adults and a meta-analysis of prospective studies
.
J Am Coll Cardiol
2015
;
66
:
1552
1562
6.
Taylor
R
,
Al-Mrabeh
A
,
Sattar
N
.
Understanding the mechanisms of reversal of type 2 diabetes
.
Lancet Diabetes Endocrinol
2019
;
7
:
726
736
7.
Sattar
N
,
Deanfield
J
,
Delles
C
.
Impact of intentional weight loss in cardiometabolic disease: what we know about timing of benefits on differing outcomes?
Cardiovasc Res
.
23 January 2023 [Epub ahead of print]. DOI: 10.1093/cvr/cvac186
8.
Sanderson
E
,
Glymour
MM
,
Holmes
MV
, et al
.
Mendelian randomization
.
Nat Rev Methods Primers
2022
;
2
:
6
9.
Sanderson
E
,
Davey Smith
G
,
Windmeijer
F
,
Bowden
J
.
An examination of multivariable Mendelian randomization in the single-sample and two-sample summary data settings
.
Int J Epidemiol
2019
;
48
:
713
727
10.
Mahajan
A
,
Spracklen
CN
,
Zhang
W
, et al.;
FinnGen
;
eMERGE Consortium
.
Multi-ancestry genetic study of type 2 diabetes highlights the power of diverse populations for discovery and translation
.
Nat Genet
2022
;
54
:
560
572
11.
Chen
J
,
Spracklen
CN
,
Marenne
G
, et al.;
Lifelines Cohort Study
;
Meta-Analysis of Glucose and Insulin-related Traits Consortium (MAGIC)
.
The trans-ancestral genomic architecture of glycemic traits
.
Nat Genet
2021
;
53
:
840
860
12.
Lagou
V
,
Mägi
R
,
Hottenga
J-J
, et al.;
Meta-Analyses of Glucose and Insulin-related traits Consortium (MAGIC)
.
Sex-dimorphic genetic effects and novel loci for fasting glucose and insulin variability
.
Nat Commun
2021
;
12
:
24
13.
Pulit
SL
,
Stoneman
C
,
Morris
AP
, et al.;
GIANT Consortium
.
Meta-analysis of genome-wide association studies for body fat distribution in 694 649 individuals of European ancestry
.
Hum Mol Genet
2019
;
28
:
166
174
14.
Evangelou
E
,
Warren
HR
,
Mosen-Ansorena
D
, et al.;
Million Veteran Program
.
Genetic analysis of over 1 million people identifies 535 new loci associated with blood pressure traits
.
Nat Genet
2018
;
50
:
1412
1425
15.
Malik
R
,
Chauhan
G
,
Traylor
M
, et al.;
AFGen Consortium
;
Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) Consortium
;
International Genomics of Blood Pressure (iGEN-BP) Consortium
;
INVENT Consortium
;
STARNET
;
BioBank Japan Cooperative Hospital Group
;
COMPASS Consortium
;
EPIC-CVD Consortium
;
EPIC-InterAct Consortium
;
International Stroke Genetics Consortium (ISGC)
;
METASTROKE Consortium
;
Neurology Working Group of the CHARGE Consortium
;
NINDS Stroke Genetics Network (SiGN)
;
UK Young Lacunar DNA Study
;
MEGASTROKE Consortium
.
Multiancestry genome-wide association study of 520,000 subjects identifies 32 loci associated with stroke and stroke subtypes
.
Nat Genet
2018
;
50
:
524
537
16.
van der Harst
P
,
Verweij
N
.
Identification of 64 novel genetic loci provides an expanded view on the genetic architecture of coronary artery disease
.
Circ Res
2018
;
122
:
433
443
17.
Shah
S
,
Henry
A
,
Roselli
C
, et al.;
Regeneron Genetics Center
.
Genome-wide association and Mendelian randomisation analysis provide insights into the pathogenesis of heart failure
.
Nat Commun
2020
;
11
:
163
18.
Nielsen
JB
,
Thorolfsdottir
RB
,
Fritsche
LG
, et al
.
Biobank-driven genomic discovery yields new insight into atrial fibrillation biology
.
Nat Genet
2018
;
50
:
1234
1239
19.
Hartiala
JA
,
Han
Y
,
Jia
Q
, et al.;
INVENT Consortium
;
CHARGE Consortium Hemostasis Working Group
;
GENIUS-CHD Consortium
;
Biobank Japan
.
Genome-wide analysis identifies novel susceptibility loci for myocardial infarction
.
Eur Heart J
2021
;
42
:
919
933
20.
Pattaro
C
,
Teumer
A
,
Gorski
M
, et al.;
ICBP Consortium
;
AGEN Consortium
;
CARDIOGRAM
;
CHARGe-Heart Failure Group
;
ECHOGen Consortium
.
Genetic associations at 53 loci highlight cell types and biological pathways relevant for kidney function
.
Nat Commun
2016
;
7
:
10023
21.
Burgess
S
,
Butterworth
AS
,
Thompson
JR
.
Beyond Mendelian randomization: how to interpret evidence of shared genetic predictors
.
J Clin Epidemiol
2016
;
69
:
208
216
22.
Hemani
G
,
Zheng
J
,
Elsworth
B
, et al
.
The MR-Base platform supports systematic causal inference across the human phenome
.
elife
2018
;
7
:
e34408
23.
Garfield
V
,
Salzmann
A
,
Burgess
S
,
Chaturvedi
N
.
A guide for selection of genetic instruments in Mendelian randomization studies of type 2 diabetes and HbA1c: toward an integrated approach
.
Diabetes
2023
;
72
:
175
183
24.
Yokoyama
H
,
Oishi
M
,
Takamura
H
, et al
.
Large-scale survey of rates of achieving targets for blood glucose, blood pressure, and lipids and prevalence of complications in type 2 diabetes (JDDM 40)
.
BMJ Open Diabetes Res Care
2016
;
4
:
e000294
25.
Litwak
L
,
Goh
S-Y
,
Hussein
Z
,
Malek
R
,
Prusty
V
,
Khamseh
ME
.
Prevalence of diabetes complications in people with type 2 diabetes mellitus and its association with baseline characteristics in the multinational A1chieve study
.
Diabetol Metab Syndr
2013
;
5
:
57
26.
Liu
B
,
Mason
AM
,
Sun
L
,
Di Angelantonio
E
,
Gill
D
,
Burgess
S
.
Genetically predicted type 2 diabetes mellitus liability, glycated hemoglobin and cardiovascular diseases: a wide-angled Mendelian randomization study
.
Genes (Basel)
2021
;
12
:
1644
27.
Patel
K
,
Zafar
MA
,
Ziganshin
BA
,
Elefteriades
JA
.
Diabetes mellitus: is it protective against aneurysm? A narrative review
.
Cardiology
2018
;
141
:
107
122
28.
Harati
H
,
Zanetti
D
,
Rao
A
, et al
.
No evidence of a causal association of type 2 diabetes and glucose metabolism with atrial fibrillation
.
Diabetologia
2019
;
62
:
800
804
29.
Zhou
JJ
,
Koska
J
,
Bahn
G
,
Reaven
P
.
Fasting glucose variation predicts microvascular risk in ACCORD and VADT
.
J Clin Endocrinol Metab
2021
;
106
:
1150
1162
30.
Welsh
C
,
Welsh
P
,
Celis-Morales
CA
, et al
.
Glycated hemoglobin, prediabetes, and the links to cardiovascular disease: data from UK Biobank
.
Diabetes Care
2020
;
43
:
440
445
31.
Chen
W
,
Wang
S
,
Lv
W
,
Pan
Y
.
Causal associations of insulin resistance with coronary artery disease and ischemic stroke: a Mendelian randomization analysis
.
BMJ Open Diabetes Res Care
2020
;
8
:
e001217
32.
Gerstein
HC
,
Bosch
J
,
Dagenais
GR
, et al.;
ORIGIN Trial Investigators
.
Basal insulin and cardiovascular and other outcomes in dysglycemia
.
N Engl J Med
2012
;
367
:
319
328
33.
Larsson
SC
,
Bäck
M
,
Rees
JMB
,
Mason
AM
,
Burgess
S
.
Body mass index and body composition in relation to 14 cardiovascular conditions in UK Biobank: a Mendelian randomization study
.
Eur Heart J
2020
;
41
:
221
226
34.
Silverwood
RJ
,
Pierce
M
,
Thomas
C
, et al.;
National Survey of Health and Development Scientific and Data Collection Teams
.
Association between younger age when first overweight and increased risk for CKD
.
J Am Soc Nephrol
2013
;
24
:
813
821
35.
Mingrone
G
,
Panunzi
S
,
De Gaetano
A
, et al
.
Metabolic surgery versus conventional medical therapy in patients with type 2 diabetes: 10-year follow-up of an open-label, single-centre, randomised controlled trial
.
Lancet
2021
;
397
:
293
304
36.
A Study of Tirzepatide (LY3298176) Compared With Dulaglutide on Major Cardiovascular Events in Participants With Type 2 Diabetes (SURPASS-CVOT), 2020
.
Accessed 11 May 2023. Available from https://clinicaltrials.gov/ct2/show/NCT04255433
37.
Dobrica
E
,
Gaman
MA
,
Cozma
MA
,
Gaman
AM
,
Diaconu
CC
.
Macrovascular and microvascular complications in type 2 diabetes mellitus: did hypertension change the setting?
J Hypertens
2019
;
37
:
e140
e141
38.
Wan
EYF
,
Fung
WT
,
Schooling
CM
, et al
.
Blood pressure and risk of cardiovascular disease in UK Biobank: a Mendelian randomization study
.
Hypertension
2021
;
77
:
367
375
39.
de Boer
IH
,
Bangalore
S
,
Benetos
A
, et al
.
Diabetes and hypertension: a position statement by the American Diabetes Association
.
Diabetes Care
2017
;
40
:
1273
1284
40.
Zheng
J
,
Zhang
Y
,
Rasheed
H
, et al
.
Trans-ethnic Mendelian randomization study reveals causal relationships between cardio-metabolic factors and chronic kidney disease
.
Int J Epidemiol
2021
;
50
:
1995
2010
41.
Ahmed
A
,
Sattar
N
,
Yaghootkar
H
.
Advancing a causal role of type 2 diabetes and its components in developing macro- and microvascular complications via genetic studies
.
Diabet Med
2022
;
39
:
e14982
42.
Guo
Y
,
Gao
J
,
Liu
Y
, et al
.
An examination of causal associations and shared risk factors for diabetes and cardiovascular diseases in the East Asian population: a Mendelian randomization study
.
Front Endocrinol (Lausanne)
2023
;
14
:
1132298
Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at https://www.diabetesjournals.org/journals/pages/license.