A total of 1,270 participants from South China enrolled in this study, starting on January 1, 2013 and ending on December 31, 2020, were divided into two groups: non-T2D group (N = 634) and T2D group (N = 636). The inclusion criteria for the case group were as follows: (1) individuals aged 18 years or older who were newly diagnosed with T2D referring to the 2012 clinical practice guidelines and its update of the American Diabetic Association (24, 25); (2) patients with T2D who had not received therapy; and (3) patients who signed the consent form. Inclusion criteria for the control group were as follows: (1) individuals aged 18 years or older, (2) in healthy conditions based on anamnesis and physical examination results, (3) no evidence of T2D or family history of T2D, and (4) patients who signed the consent form. The non-T2D participants, got a health examination at the same hospital during the same period, and were enrolled as the control group. Subjects were excluded from enrollment in the study upon presenting (1) a history of hypertension (HTN), CAD, HF, and stroke. However, the history of lacunar cerebral infarcts (LCIs) or (and) atrial fibrillation (AF) at enrollment, the new onset of HTN, CAD, HF, and stroke during the follow-up will be not considered as exclusion criteria. The subjects who had a history of LCI or/and AF but no history of HTN, CAD, HF, and stroke at the time of enrollment were not considered as exclusion criteria. Stroke and LCI were evaluated by magnetic resonance imaging (MRI) or (and) computed tomography (CT). Stroke was evaluated at the time of enrollment by MRI or (and) CT; (2) high levels of aspartate aminotransferase (AST) and alanine transaminase (ALT) (more than 3 times the upper limit of the normal); (3) low estimated glomerular filtration rate (less than 90 ml/min•1.73 m²); (4) or (and) any other factors that contribute to serum lipid abnormalities, such as medical disorders (thyroid disease, Cushing’s syndrome, and pheochromocytoma) and drugs [thiazide diuretics, beta-receptor blockers (BBs), steroid, estrogen, progesterone, immunosuppressants, or retinoids]. Dyslipidemia is defined as an increase in triglyceride levels (TIRG) ≥ 1.7 mmol/L, total cholesterol (TC) ≥ 4.0 mmol/L, low-density lipoprotein cholesterol (LDL-C) ≥ 1.4 mmol/L, high-density lipoprotein cholesterol (HDL-C) < 1.0 mmol/L, apolipoprotein B (ApoB) ≥ 80 mg/dL, apolipoprotein A-I (ApoA-I) < 120 mg/dL or lipoprotein (a) [Lp(a) ≥ 300 mg/dL], and any their combinations. All patients with dyslipidemia were newly diagnosed at the time of enrollment in this study. Dyslipidemia diagnosis is in accordance with the recommendations on lipid disorders management of ESC/EAS (26). Echocardiography is evaluated using the standard sections at enrollment according to the recommendations of the American Society of Echocardiography (26). Blood biochemistry analyses were carried out at the time of enrollment according to standard analytical methods.

In this study, the primary endpoints are defined as the risks of new-onset/recurrent acute coronary syndrome (ACS), new-onset of stroke, new-onset of HF, and new-onset of AF during the follow-up. The secondary endpoints are defined as the risks of LCI at enrollment, AF at enrollment, and the total AF risk at the end of the follow-up period. These endpoint events were identified from three information sources, including the Center for Disease Control and Prevention, medical records, and participants and their families. Subjects were enrolled in the study on their first day of T2D diagnosis. December 31, 2020 was the final follow-up date. During the follow-up, events were diagnosed with reference to the following guidelines: (1) new-onset/recurrent ACS is defined as ST-elevation of myocardial infarction (MI), non-ST-elevation of MI or unstable angina, confirmed by cardiac biomarkers and coronary angiography. (2) LCIs are small infarcts (2–20 mm in diameter) caused by the occlusion of small penetrating end-arteries. New-onset of stroke is defined as the presence of a focal/global cerebral event including hemorrhagic stroke (HS), ischemic stroke (IS), and cardiogenic stroke (CS). Both LCI and strokes were evaluated by MRI or (and) CT at enrollment. (3) New-onset of HF is defined as the first occurrence of a HF admission (including at least one overnight hospitalization to emergency) (27), including HF with reduced ejection fraction [HFrEF, left ventricular ejection fraction (LVEF) ≤ 40%], HF with mildly reduced ejection fraction (HFmrEF, LVEF = 41–49%), and HF with preserved ejection fraction (HFpEF, LVEF ≥ 50%). (4) The AF is assessed according to prior medical records or electrocardiogram records from the enrollment to follow-up, including AF at enrollment and new-onset of AF. The new-onset of AF is defined as the newly detected AF in participants with no previous diagnosis of this condition before enrollment.

Nine KATP SNPs (rs2285676, rs11046182, rs1799858, rs4148671, rs78148713, rs145456027, rs147265929, rs61928479, and rs141294036) were identified using the MassARRAY (Sequenom) system as previously described (26). Specific primers for the 9 KATP SNPs are shown in Supplementary Table 1. The 9 KATP SNPs have 100% accurate genotypes for further analyses.

All analyses were performed using PASS 15.0 (Statistical Solution Ltd, Cork, Ireland) and SPSS 20.0 (SPSS, Chicago, IL). LDlink (LDmatrix) online tool¹ was used to calculate the linkage disequilibrium (LD) for the 9 KATP variants in the Chinese populations as shown in Supplementary Figure 1. The Hardy-Weinberg equilibrium was investigated for both non-diabetic and diabetic subjects as shown in Supplementary Table 2. Data of categorical variables are presented as numbers (percentages) while continuous variables are presented as mean ± standard deviation. The difference among the subjects without or with T2D was analyzed by χ²-test for the categorical variables, or by independent sample t-test for the continuous variables. The dominant models for the minor allele of the 9 KATP SNPs were tested to obtain odds ratios (ORs)/hazard ratios (HRs), 95% confidence intervals (CIs), and P-values. Associations of the 9 KATP SNPs with T2D were assessed using binary logistic regression model as well as LCI, different types of dyslipidemia, and low-grade inflammation levels [high-sensitivity C-reactive protein (HsCRP) ≥ 3.0 mg/L]. Associations of the 9 KATP SNPs with new-onset/recurrent ACS were assessed using Cox proportional hazards regression model; the new-onset of stroke, HF, and AF were assessed similar to ACS. In particular, the relationships between KATP variants and different stroke subtypes (HS, IS and CS) were assessed using multinomial logistic regression model; the different HF subtypes (HFpEF, HFmrEF, and HFrEF) were assessed similarly. Bonferroni correction was conducted to protect from Type I error. Further, the heritability (explained variance,%) of KATP variants for events was calculated according to the method reported by Yang et al. (28). All probabilities are two-tailed. A P-value < 0.05 is considered statistically significant.

Among all the study subjects (N = 1,270), 4.1% were presented with a history of LCI (52 patients), and 3.5% with a history of AF (44 patients) at the time of enrollment, as shown in Table 1. Subjects with or without T2D at enrollment indicate obvious differences in LCI incidence rates (P = 0.024), levels of systolic blood pressure (SBP, P < 0.001) and diastolic blood pressure (DBP, P < 0.001), white blood cell count (WBC, P < 0.001), serum levels of TRIG (P < 0.001), HDL-C (P < 0.001), ApoA-I (P = 0.018), uric acid (UA, P < 0.001), HsCRP (P = 0.030), left atrial end-diastolic dimension (LAD, P = 0.012), LVEF (P = 0.028), left ventricular mass index (LVMI, P = 0.001) and E/e’ ratio (P = 0.003). During the follow-up period, none of the study subjects in the control group progressed to T2D, none dropped out of the study, and the subjects suffered from cardiovascular conditions are as follows: new-onset/recurrent ACS (416 patients, 32.8%), new-onset of stroke (58 patients, 4.7%), new-onset of HF (282 patients, 22.2%) and new-onset of AF (94 patients, 7.4%), as shown in Supplementary Table 3. Subjects with or without T2D present obvious differences for new-onset of HF (P = 0.005) and HF subtypes (P = 0.013); the differences remain significant upon factoring the medications in the analyses, such as mineralocorticoid receptor antagonist (MRA, P = 0.047), BBs (P = 0.015), and renin-angiotensin system inhibitors (RSIs, P < 0.001) at the end of follow-up.

As shown in Figure 1 and Supplementary Table 4, KATP SNPs rs2285676 (AA + GA, adjusted OR = 1.43, 95% CI: 1.13–1.81, P = 0.003), rs1799858 (CC, adjusted OR = 1.42, 95% CI: 1.12–1.78, P = 0.004), rs145456027 (TT, adjusted OR = 2.94, 95% CI: 1.39–6.22, P = 0.005), rs61928479 (AA, adjusted OR = 1.71 95% CI: 1.18–2.47, P = 0.005), and rs141294036 (CC, adjusted OR = 1.45, 95% CI: 1.15–1.83, P = 0.002) significantly correlate with moderate to high risk T2D except for rs11046182, rs4148671, rs78148713, and rs147265929 (all adjusted P > 0.05).

As shown in Figure 2, Supplementary Table 5, and Supplementary Figure 2, KATP rs2285676 (AA + GA, adjusted HR = 1.37, 95% CI: 1.10–1.70, P = 0.005) and rs141294036 (TT + CT, adjusted HR = 1.59, 95% CI: 1.28–1.99, P < 0.001) are associated with a moderate risk of new-onset/recurrent ACS after a median follow-up of 45.8-months (IQR: 27.4–55.3 months).

As shown in Figure 3A, Supplementary Table 6, and Supplementary Figure 3, KATP rs1799858 (CC, adjusted HR = 2.58, 95% CI: 1.22–5.43, P = 0.013) and rs141294036 (CC, adjusted HR = 2.30, 95% CI: 1.16–4.55, P = 0.017) correlate with increased risk of new-onset stroke based on the median follow-up period of 50.8-months (IQR: 44.0–58.7 months). As shown in Figure 3B and Supplementary Table 7, further analyses of stroke subtype indicate that rs1799858 (CC, adjusted OR = 2.38, 95% CI: 1.11–5.10, P = 0.025) and rs141294036 (CC, adjusted OR = 1.95, 95% CI: 1.04–3.66, P = 0.037) are associated only with IS.

As shown in Figure 4A, Supplementary Table 8, and Supplementary Figure 4, KATP rs1799858 (TT + CT, adjusted HR = 2.78, 95% CI: 2.07–3.74, P < 0.001) and rs141294036 (TT + CT, adjusted HR = 1.45, 95% CI: 1.07–1.96, P = 0.015) are associated with high HF risk based on a median follow-up of 48.1-months (IQR: 39.1–56.8 months). As shown in Figure 4B and Supplementary Table 9, further analyses of HF subtypes show that rs1799858 is associated only with high HFpEF risk (TT + CT, adjusted OR = 3.46, 95% CI: 2.31–5.18, P < 0.001) while rs141294036 is associated with high HFmrEF risk (TT + CT, adjusted OR = 2.74, 95% CI: 1.05–7.15, P = 0.039).

As shown in Supplementary Table 10, two KATP variants are associated with high AF risk at enrollment as follows: rs1799858 (TT + CT, adjusted OR = 2.15, 95% CI: 1.06–4.39, P = 0.035) and rs141294036 (CC, adjusted OR = 2.42, 95% CI: 1.15–5.07, P = 0.020). As shown in Figure 5, Supplementary Table 11, and Supplementary Figure 5, rs1799858 (TT + CT, adjusted HR = 2.05, 95% CI: 1.25–3.37, P = 0.004) and rs141294036 (CC, adjusted HR = 2.31, 95% CI: 1.40–3.82, P = 0.001) are associated with high risk of new-onset of AF based on a median follow-up of 50.1-months (IQR: 42.7–58.4 months).

As shown in Figures 6A–E and Supplementary Tables 14–18, KATP rs2285676 (AA + GA genotype) is associated with the elevated risk of higher TRIG serum levels (adjusted OR = 1.37, 95% CI: 1.02–1.84, P = 0.035), LDL-C (adjusted OR = 1.82, 95% CI: 1.13–2.92, P = 0.014), ApoB (adjusted OR = 1.34, 95% CI: 1.03–1.74, P = 0.031), and Lp(a) (adjusted OR = 1.48, 95% CI: 1.13–1.95, P = 0.004). The rs1799858 (TT + CT) is associated with a higher risk of elevated LDL-C serum levels (adjusted OR = 2.26, 95% CI: 1.39–3.69, P = 0.001) and lower ApoA-I serum levels (adjusted OR = 1.57, 95% CI: 1.13–2.19, P = 0.008) while CC genotype is associated with higher Lp(a) serum levels (adjusted OR = 1.52, 95% CI: 1.15–2.01, P = 0.003). Similarly, rs141294036 (TT + CT) is associated with the elevated risk of lower serum ApoA-I levels (adjusted OR = 1.72, 95% CI: 1.28–2.30, P < 0.001) while CC genotype is associated with higher Lp(a) serum levels (adjusted OR = 1.56, 95% CI: 1.19–2.03, P = 0.001). KATP rs4148671 (TT + CT) only correlates with increased risk of higher serum TRIG levels (adjusted OR = 1.62, 95% CI: 1.01–2.60, P = 0.045) while rs61928479 (AA) correlates with the elevated risk of lower serum ApoA-I levels only (adjusted OR = 2.05, 95% CI: 1.22–3.46, P = 0.007). Three KATP SNPs (rs11046182, rs78148713 and rs145456027) do not associate with any type of lipid disorder [HDL-C, ApoB, ApoA-I, and Lp(a), all P > 0.05]. As shown in Figure 6F and Supplementary Table 21, KATP rs1799858 (TT + CT, adjusted OR = 1.47, 95% CI: 1.10–1.96, P = 0.009), rs2285676 (AA + GA, adjusted OR = 1.42, 95% CI: 1.05–1.91, P = 0.023), and rs141294036 (CC, adjusted OR = 1.49, 95% CI: 1.11–2.00, P = 0.008) correlate with the increased risk of higher serum HsCRP levels.

This study demonstrates that the Chinese participants with the genotypes of rs2285676 (AA + GA), rs1799858 (CC), and rs141294036 (TT + CT) have a 42–45% increased risk of T2D, and the total heritability of the 3 KATP variants for T2D risk is 2.1% (Supplementary Table 4). These findings are in controversy with the previous studies by Baier et al. (14) who reported no association between KATP rs2285676 and T2D in specific Amerindian-derived population, and by Odgerel et al. (13) who reported an association between the T-allele of rs1799858 and T2D in diabetic Mongolians rather than the CC genotype. The subjects with high diabetic-risk genotypes with different KATP variants correlate with increased T2D risk, but the effects show the characteristics of ethnic heterogeneity as follows: The French (rs1799859/AA + AG, by 69%) (6), the Turks (rs1799859/AA + AG, by 3.8-fold) (7), the Finns (rs1799859/AA + AG, by 2-fold), the Japanese (rs5219/KK + EK, by 32%) (12), the Mongolians (TT + CT of rs1799858 and rs2074308, approximately by 80%) (13) and the Pima Indians (chr11:17417205/T-allele, by 1-fold) (14). Two meta-analysis studies confirm that KATP mutations correlate with increased T2D risk (about by 15–16%) in both the Caucasians (e.g., the Europeans) and the non-Caucasians (e.g., the Japanese) (4, 5). Furthermore, KATP variants are associated with antidiabetic efficacy of sulfonylurea hypoglycemic agents (e.g., gliclazide) as well as non-sulfonylureas (e.g., repaglinide), especially secondary sulfonylurea failure (29). This phenomenon is also seen in the Chinese Han subjects with T2D (e.g., rs1801261) (30). These findings suggest that the 3 KATP variants can be targeted to evaluate the effectiveness of hypoglycemic medications in the Chinese in addition to being latent genetic predisposition markers for T2D risk.

Currently, the mechanism by which KATP variants induce high diabetes risk remains unclear, possibly involving inadequate insulin secretion, resistance to insulin, or (and) impaired glucagon secretion. First, the KATP mutations are associated with impaired glucose-induced insulin secretion in the Turks (e.g., rs1799854, rs1799859, and rs5219) (7), the Finns (e.g., rs5219) (31), and the French-Canadians (e.g., rs2285676 and rs11024273) (32). Secondly, KATP variants associate with decreased insulin sensitivity in the Poles (e.g., rs80356622 and rs80356624) (33), the Finns and the Americans (e.g., rs5219) (31). Lastly, KATP rs5219 correlates with impaired glucose-induced suppression of glucagon secretion (34). What is more important is that both animal and clinical studies indicate that KATP variants manifest their deleterious effects at early stages in the progression of T2D (e.g., during the progression from normal blood glucose levels to impaired glucose tolerance) (35), especially under the influence of environmental factors, such as diet (e.g., rs5219 and alcohol consumption in the Koreans) (36) and lifestyle (e.g., rs3758947 and physical inactivity in the Finns) (37). Indeed, R1420H (chr11:17417205C > T) correlates with increased risk of T2D (by 2-fold) with a 7-year earlier onset age (14). A recent study shows that the Chinese Han subjects with KATP rs5219 orrs1801261 are more susceptible to the development of prediabetes (38).

Arteriosclerosis cardiovascular disease, such as ACS and stroke are the major macro-vascular complications in patients with T2D (39), and are the leading cause of increased mortality among diabetics, even in people who optimally control the common risk factors for ASCVD. Among the patients with ACS, those with T2D are particularly at an increased risk for recurrent cardiovascular events and premature death, especially in China (40). Our results show that the Chinese participants with high diabetic risk genotype (A-allele) of rs2285676 present 37% increased the risk of new-onset/recurrent ACS after a median follow-up of 45.8-months. The T-allele of rs141294036 is not a diabetic risk predisposition allele, but the Chinese subjects with this mutation are at the highest risk for new-onset/recurrent ACS (increased by 59%). On the other hand, the new-onset of stroke risk increases by 1.58-fold and 1.30-fold among the Chinese subjects expressing the high diabetic risk genotype (CC) of rs1799858 and rs141294036, respectively, after a median follow-up of 50.8-months. However, the CC genotype of rs1799858 and rs141294036 correlates only with moderate and high risk of IS. In addition to being related to the occurrence risk, a recent study further showed that the KATP variants (e.g., rs2237982, rs2283261, rs3819521, and rs8192695) are also linked to the progression of vascular events, and are involved in abnormal transcription in the target-tissue and promoter/enhancer markers (41). Compared to the heritabilityofrs141294036 for new-onset stroke risk (0.44%, Supplementary Table 6), the heritability of the loci for new-onset/recurrent ACS risk (1.34%, Supplementary Table 5) are much higher. These results suggest that the genotypes, rs2285676 (AA + AG), rs1799858 (CC), and rs141294036 (CC) are optimal genetic markers for macro-vascular events in diabetic patients, especially rs141294036 for a new-onset/recurrent ACS risk.

Type 2 diabetes is a very important independent risk factor for HF. Previous studies show that KATP rs5219 prevalence is increased in the American patients with HF (21) while 3 KATP variants are associated with left ventricular remodeling in ischemic HF subjects in Ukraine (e.g., rs5215, rs5219 and rs757110), suggesting that the KATP mutations may correlate with increased HF risk. This study shows a 1.78-fold increase in HF among the Chinese subjects with TT + CT genotype of rs1799858 after a median follow-up of 48.1-months. Compared to the association between rs1799858 and HF, rs141293036 (TT + CT) correlates with weaker HF incidence risk (only increased by 45%). The heritability of HF is heterogeneous, and the genetic characteristics of different HF subtypes are still poorly understood. Further analysis of HF subtypes first indicates that rs1799858 (TT + CT) only correlates with a high risk of HFpEF while rs141294036 (TT + CT) correlates with a moderate risk of HFmrEF. The heritability of rs1799858 for a new-onset of HF risk is up to 3.48% (Supplementary Table 8), especially 2.78% for HFpEF (Supplementary Table 9). As far as the current HFpEF and HFmrEF prevention and treatment strategy are concerned, the “phenotype”- based targeted therapy is a more feasible priority strategy (42). Hence, these findings indicate that the TT + CT genotype of rs1799858 and rs141294036 can serve as a potential phenotypic marker for HFpEF and HFmrEF phenotypes, especially, rs1799858.

Type 2 diabetes is also one of the strongest independent risk factors for AF. However, the relationships between KATP variants and AF risk in the Chinese diabetic subjects remain unclear. This study shows that the Chinese subjects with high diabetic risk genotype (CC) of rs141294036 are at 1.42-fold more risk of having AF at the time of enrollment. The T-allele of rs1799858 does not correlate with diabetes risk, but the Chinese subjects with the mutation are at a higher risk of AF (approximately increased by 1.15-fold). After a median follow-up of 50.1-months, a similar association between the two variants and the risk of a new-onset of AF was found, where the AF risk was more than doubled in the presence of any of these variants (Supplementary Table 11). In this study, the total heritability of rs1799858 and rs141294036 for total AF risk reaches up to 1.85% (Supplementary Table 12). Similarly, an isolated study indicates that the KATP variant, rs387906805 correlates with AF originating from the vein of Marshall (43). A large sample study also indicates an association of KATP rs72554071 with increased AF susceptibility among the Americans based on 8-years follow-up (22). The pathogenic mechanism involves shortening of the action potential duration mediated by a transient outward current (22). These results suggest that rs141294036 (CC) may be a preferred predisposition marker for diabetic AF in the Chinese subjects.

Glucose and lipid metabolism disorders, along with chronic low-grade inflammation, are essential in the initiation and progression of diabetic cardiovascular events. Our study and many other studies, demonstrates the association of KATP variants with glucose metabolism disorders; however, lipid metabolism disorders and inflammation are rarely highlighted. This study indicates that the three diabetic-risk related KATP variants are indeed correlated with dyslipidemia but exhibited heterogeneity as follows: none of the 3 KATP variants correlates with increased serum TC levels (Supplementary Table 19), and decreased serum HDL-L level (Supplementary Table 20). KATP rs2285676 (A-allele) correlates with a higher risk of all subtypes [e.g., TRIG, LDL-C, ApoB and Lp(a)] of dyslipidemia except HDL-C and ApoA-I. The high diabetic risk genotype (CC) of rs1799858 and rs141294036only correlates with moderate risk of increased serum Lp(a) levels while its counterpart genotype (TT + CT) correlates with a moderate risk of decreased serum ApoA-I levels. Besides ApoA-I, the KATP rs1799858 (TT + CT) correlates with a high risk of increased serum LDL-C levels. On the other hand, the three diabetic-risk related KATP variants are associated with a moderate risk of low-grade inflammation levels (HsCRP ≥ 3.0 mg/L).

The clinical heterogeneity of diabetic cardiovascular phenotypes may relate to the combinations of hyperglycemia, different subtypes of dyslipidemia, and chronic low-grade inflammation (44). First, the CC genotype of rs1799858 and rs141294036 correlate with a high IS risk. These two variants correlate with a high risk of higher serum Lp(a) levels. In addition, two recent large sample studies confirm that high serum concentration of Lp(a) is associated with an increased risk of IS, both observationally and causally based on human genetics (45, 46). Second, the TT + CT genotype of rs1799858 and rs141294036 correlate with HFpEF and HFmrEF risks. Both the variants correlate with a high risk of decreased serum in the ApoA-I level. Among the lipoprotein components related to HDL-C, ApoA-I presents the greatest independent risk factor for cardiovascular risk events and is a better predictive marker for HF (47). In mouse models of HF, a reconstituted HDL-C improves the outcomes, especially HFpEF (48). Importantly, under the same adverse effect of HF incidence rate (e.g., 10%), HF occurs earlier (about 1 year; Supplementary Figure 4) in subjects with TT + CT genotype of rs1799858 compared to those with TT + CT genotype of rs141294036. Besides the high risk of decreased serum ApoA-I level, rs1799858 is also related to the coexisting elevated risk of high serum levels of LDL-C and HsCRP. Third, rs2285676 (AA + AG) and rs141294036 (TT + CT) are associated with ACS risk. The difference between them is that the former almost correlates with a higher risk of all subtypes of dyslipidemia except HDL-C and ApoA-I, where the latter shows the complete opposite. A consensus is established on the relationship between arteriosclerotic cholesterol [e.g., TRIG, LDL-C, ApoB and Lp(a)] and ACS risk, but little is known about its relationship with ApoA-I. Lower serum ApoA-I levels are associated with increased ACS risk and is manifested in individuals with a high cardiovascular risk (49). However, coronary atherosclerosis is not reduced by ApoA-I mimetic infusions in patients with ACS and high plaque burden (50). Lastly, rs1799858 (TT + CT) and rs141294036 (CC) are associated with AF risk. Both of them are associated with a high risk of low-grade inflammation (HsCRP ≥ 3.0 mg/L). A recent study shows that chronic low-grade inflammation correlates with increased AF risk in the Chinese participants having a metabolic syndrome (51).

This study is the first long follow-up assessment of the genotype-cardiovascular phenotype profiles of KATP variants among diabetic subjects in China. Several limitations worth mentioning are as follows: First, the study determines the above associations successfully, but the direct evidence with regards to its pathophysiology is yet to be investigated. Second, two potential factors may influence the associations, including the sample size and cardiovascular events (e.g., stroke and HF) during the follow-up. Importantly, it is necessary to explore the association of glycemic parameters, such as HbA1C and FBG with the new onset of cardiovascular phenotypes, given there is a strong association between the glycemic control and macro-vascular complications. Initially, the current work only demonstrates the association between KATP and diabetic cardiovascular phenotypes. These results and plans should be further validated by large-scale, prospective cohort studies with large-scale sample size. Third, there is still a risk of potential Type I error so that Bonferroni adjustment is carried out to correct the significant thresholds. Fourth, microvascular dysfunction (MVD) in T2D is present in many tissues or organs and it manifests itself as a spectrum of specific phenotypes: LCI in the brain (52), HFpEF in the heart (53), and microalbuminuria in the kidney (54). We observed a potential clinical association between KATP genetic variants and MVD-related phenotypes (e.g., LCI and HFpEF), but the direct evidence based on MVD as the pathophysiological entity is still lacking and needs further investigation. Lastly, the crosstalk between the environmental and genetic factors results in the development of diabetes and its associated CVDs. Additionally, the effect of non-coding RNA on the pathophysiological processes from elevated serum glucose levels to different diabetic cardiovascular phenotypes needs further investigation.