The association between hypoglycemia and panvascular disease in type 2 diabetes patients: The correlation structure mediated by inflammatory and metabolic abnormalities

Abstract

Background

Hypoglycemia in type 2 diabetes patients is associated with metabolic abnormalities and inflammation that affect vascular beds. The relationship between hypoglycemia and panvascular disease (PVD) is not clear. This study aimed to investigate the association between PVD and hypoglycemia, and to identify potential mediators.

Method

This retrospective cross-sectional study enrolled Patients from two centers in Chongqing China, and the results were further validated using UK Biobank data. Logistic regression was used to test the association of hypoglycemia and PVD. Stratification and interaction analyses to test the effects across study subgroups. Forward (hypoglycemia to PVD) and reverse (PVD to hypoglycemia) relationships were analyzed by structural equation modeling (SEM), which included interleukin-6, neutrophil-to-lymphocyte ratio, uric acid, hemoglobin A1c, and systolic blood pressure.

Results

22,128 patients diagnosed with T2DM at two large centers and 44442 T2DM participants from the UK Biobank were enrolled. A significant association between hypoglycemia and PVD was found. Subgroup analysis found hypoglycemia was more strongly associated with PVD in patients with inflammatory abnormalities and metabolic dysfunction. SEM suggested a correlation structure between hypoglycemia and PVD which might mutually aggravated each other through inflammation and metabolism pathways.

Conclusion

This is the first study that described the correlation structure between hypoglycemia and PVD with a large population. Within this potential mutual association, inflammation and metabolism might be mediators. Our study also highlights the insufficient attention clinicians pay to hypoglycemia and PVD, and further attention is needed in future clinical practice and research.

Keywords

panvascular disease (PVD)hypoglycemia type 2 diabetes (T2DM)inflammation metabolism correlation structure

Introduction

Diabetes is one of the most common noncommunicable diseases. It affects 422 million people worldwide and can lead to various irreversible complications, with the vascular system being a primary target.^1–3 Hypoglycemia is a common occurrence in diabetes.^1,2 A previous study reported that 44.7% of type 2 diabetes mellitus (T2DM) patients experienced minor hypoglycemic events, with an incidence of 52.0% in the intensive treatment group and 37.3% in the standard treatment group.³

Hypoglycemia has traditionally been considered as the hazardous side effect of diabetes treatment.⁴ However, recent studies also held the position that occurrence of hypoglycemia may not only be a result of blood glucose control issues but also indicate the overall vulnerability of the patient’s health, representing a worse metabolic and inflammatory condition.^1,5–10 Besides, hypoglycemia is reported to be associated with an increased risk of various vascular diseases in diabetes patients.^1–3 Previous studies have suggested that the occurrence of hypoglycemia increases the increased risk of vascular diseases, including a prothrombotic state, stroke, myocardial infarction, microvascular complications, and major adverse cardiovascular events in diabetes patients.^3–5 Unfortunately, these previous studies focused on the association between hypoglycemia and a single vascular bed rather than on considering vasculature as a whole^11–13 even though previous research has already pointed out that panvascular disease (PVD) is associated with greater challenges in disease management, increased treatment complexity, and poorer clinical prognosis in diabetes patients.^14–16

Some studies have already pointed out that metabolic changes associated with hypoglycemia in diabetes may stimulate excessive production of reactive oxygen species and the release of various pro-inflammatory factors that are involved in the pathogenesis of vascular disease.^1,5–10 However, the mechanism mediated under the association between hypoglycemia and vascular disease still needs further investigation, especially for PVD.

Therefore, our study aimed to investigate the association between PVD and hypoglycemia, and to identify potential inflammatory biomarkers that may mediated metabolic this relationship. Besides, our study also sought to highlight the clinical and research importance of hypoglycemia and panvascular disease, both of which warrant greater attention.

Methods

Study centers and patient population

This two-center retrospective observational study enrolled T2DM patients who were admitted to The First Hospital of Chongqing Medical University from June 2022 to May 2024 and The First Affiliated Hospital of Chongqing University of Chinese Medicine from April 2019 to May 2024 (Figure 1).

Figure 1.

The graphical abstract.

The inclusion criteria were: (1) ≥18 years of age (2) diagnosed with T2DM (3) Patients signed a consent form upon admission and agreed to the use of their medical data for medical research purposes; The exclusion criteria: (1) Patients who were pregnant or had diabetes other than type 2 (2) Patients who lacked eligible blood glucose monitoring data.

Collection of clinical data

Physicians at both centers diagnosed the patients according to the International Classification of Diseases (ICD)-10 codes. The codes used are shown in the Supplementary Methods. All other patient data including medication information were retrieved from the electronic medical records system. Python is used for data filtration and management of the data are extracted data.

The data included in the analysis were the patient’s unique identification number, blood glucose monitoring records, the discharge diagnosis, laboratory test and imaging results, patient clinical and demographic characteristics, and other related information. Missing data were addressed using a predefined rule based on the proportion of missingness. Variables with <20% missing values were imputed using multivariate multiple imputation to minimize information loss. Variables with ≥20% missingness were analyzed using complete-case analysis to avoid introducing bias from excessive imputation. The primary analyses in this study were conducted using the imputed dataset, whereas complete-case analysis was applied only to selected SEM models and stratified analysis (Supplemental Figures 3 and 4). The sample size for each analysis was indicated at the end of the figures. Sensitivity analyses comparing imputed and complete-case datasets yielded consistent results. Test results that were continuous variables had corresponding categorical reference variables in the medical record system. Both continuous and categorical forms of the test results were collected. In case of multiple test results, the average value was used for continuous variable analysis. IL-6 has more than one reference value because more than one assay method was used. The reference values are in the Supplementary Methods.

PVD and hypoglycemia

During hospitalization, patients at both centers with glucose monitors had least one capillary blood or venous blood glucose every day in the morning and at any other time that hypoglycemic symptoms occurred or when the doctor deems it necessary. According to previous research and guideline, hypoglycemia in this study was defined as any blood glucose monitoring result ≤3.9 mmol/L.^4,17 As described in previous studies, expert consensus, and clinical practice, PVD is defined as the presence of vascular lesions across multiple vascular beds, mainly manifesting as target organ damage.^{11,14–16,18} Therefore, our study divided the vessel beds into six parts: brain vessel bed, retinal vessel bed, renal vessel bed, heart vessel bed, peripheral (limb) vessels bed, the aorta and it’s major branches (Figure 1). According to previous consensus and research, the diagnoses of lesions in these vessel beds or related targeted organ damage was used to identify vascular diseases and patients with two or more involved vessel beds were identified as PVD cases.^{11,14–16,18} The diagnoses for different vessel beds are shown in the Supplementary Methods.

Validation using data from the UK biobank

UK Biobank data was incorporated into this study to validate the relationship between hypoglycemia and PVD in diabetes patients (Figure 1). The study was supported by the UK Biobank (ID 387995) and approved by the Northwest Multi-Centre Research Ethics Committee (Ref 11/NW/0382 on June 17, 2011). All the participants provided written informed consent at the initial evaluation in the UK Biobank study. Details About the method of UK biobank data validation was shown in supplementary methods.

SEM

Using the double center data, SEM were applied to assess the possible forward (hypoglycemia to PVD) and reverse (PVD to hypoglycemia) correlation structure between hypoglycemia and PVD. Based on data from previous studies, interleukin (IL)-6 and the neutrophil-to-lymphocyte ratio (NLR) were selected as inflammatory markers and systolic blood pressure (SBP), uric acid (UA), and HbA1c were chosen as metabolic markers. These variables were incorporated into the SEMs to explore the mediating roles of inflammation and metabolism in this relationship.^{5,6,8,19–21}

Statistical analysis

Continuous variables were reported as means ± standard deviation or medians and interquartile range and compared by analysis of variance (ANOVA) or the Kruskal–Wallis test, as appropriate. Categorical variables were reported as frequencies and percentages and compared using Pearson’s chi-squared test or Fisher’s exact test as appropriate. Univariate logistic regression was used to identify clinical characteristics associated with vascular lesions and evaluate their correlation with PVD. Variables with a significant association were then included in a multivariate logistic regression model to test the association of hypoglycemia and PVD. Additional indicators associated with hypoglycemia or PVD were subsequently included in multivariable logistic regression adjustment model for sensitivity analyses. For some categorical variables, subgroup analyses using multivariable logistic regression and interaction testing of likelihood ratio were performed. Effect sizes were reported as odds ratios (ORs) with their 95% confidence intervals (CIs). Structural equation modeling (SEM) was used to examine the direct and indirect bidirectional forward (hypoglycemia on PVD) and reverse (PVD on hypoglycemia) association with hypoglycemia, with a focus on the mediating roles of inflammation and metabolism in these relationships. The SEM results are reported as estimated path coefficients (E).^12,13 The statistical analysis and figure drawing were performed with R version 4.4.2 (https://www.R-project.org, The R Foundation for Statistical Computing) or Python 3.13.0 (https://www.python.org, Python Software Foundation). P < 0.05 indicated statistical significance. Variables with <30% missing data were imputed using the multiple imputation by chained equations method.

Results

Clinical characteristics

A total of 22,128 patients who were diagnosed with T2DM at two large centers met the inclusion criteria (Figures 1 and 2). For the combined data (Table 1), the prevalence of hypoglycemia was 16.7% (n = 1099) in the PVD group and 11.1% (n = 1722) in the non-PVD group. The median age was 65.7 (IQR: 56.0, 73.0) years for the non-PVD group and 71.0 (IQR: 62.0, 78.0) years for PVD group. The median body mass index (BMI) was 24.4 (IQR: 22.6, 26.4) kg/m² for the non-PVD group and 24.2 (IQR: 22.6, 26.0) kg/m² for PVD group. The PVD group included 41.5% women compared to 45.4% in the non-PVD group.

Figure 2.

Study flow diagram.Abbreviations: PVD, panvascular disease; HCUCM, The First Affiliated Hospital of Chongqing University of Chinese Medicine; HCQMU, The First Affiliated Hospital of Chongqing Medical University.

Table 1.

Clinical characteristics of patients with and without PVD in both study centers.

Variables	Non-PVD	PVD	p
N	15553	6575
Hypoglycemia, (%)	1722 (11.1)	1099 (16.7)	<0.01
Age (Years), median [IQR]	65.7 [56.0, 73.0]	71.0 [62.0, 78.0]	<0.01
Female, (%)	7068 (45.4)	2726 (41.5)	<0.01
Smoke (yes), (%)	4438 (28.5)	2366 (36.0)	<0.01
Alcohol use (yes), (%)	3011 (19.4)	1550 (23.6)	<0.01
BMI (kg/m²), median [IQR]	24.4 [22.6, 26.4]	24.2 [22.6, 26.0]	<0.01
Hypertension (yes) (%)	8904 (57.2)	5243 (79.7)	<0.01
HbAlc (%), median [IQR]	7.5 [6.6, 8.7]	7.6 [6.7, 8.9]	<0.01
TC (mmol/L), median [IQR]	4.2 [3.6, 4.8]	4.0 [3.4, 4.6]	<0.01
TG (mmol/L), median [IQR]	1.6 [1.2, 2.2]	1.6 [1.2, 2.2]	<0.01
HDL (mmol/L), median [IQR]	1.1 [0.9, 1.3]	1.1 [0.9, 1.3]	<0.01
LDL (mmol/L), median [IQR]	2.4 [1.9, 2.9]	2.2 [1.7, 2.7]	<0.01
SBP (mmHg), median [IQR]	135.4 [123.0, 148.4]	140.0 [128.0, 151.0]	<0.01
DBP (mmHg), median [IQR]	79.0 [72.0, 87.0]	77.6 [70.8, 85.0]	<0.01
Statin (%)	5483 (35.3)	4843 (73.7)	<0.01
HLP (yes), (%)	5730 (36.8)	2672 (40.6)	<0.01
Center (HCUCM)	2542 (16.3)	2554 (38.8)	<0.01
Insulin using (yes), (%)	10352 (66.6)	5178 (78.8)	<0.01
Insulin pump using (yes), (%)	1200 (7.7)	1683 (25.6)	<0.01
NLR, median [IQR]	3.4 [2.3, 5.2]	3.7 [2.6, 5.3]	<0.01
CTMT (yes), (%)	2568 (16.5)	2568 (39.1)	<0.01
Brain vascular disease, (%)	3136 (20.2)	4081 (62.1)	<0.01
Retinal vascular disease, (%)	278 (1.8)	1484 (22.6)	<0.01
Heart vascular disease, (%)	2453 (15.8)	3882 (59.0)	<0.01
Peripheral vascular disease, (%)	588 (3.8)	2740 (41.7)	<0.01
Aorta and major branches, (%)	274 (1.8)	1293 (19.7)	<0.01
Renal vascular disease, (%)	848 (5.5)	3227 (49.1)	<0.01

Abbreviations: BMI, body mass index; HbA1c, hemoglobin A1c; TC, total cholesterol; TG, triglycerides; HDL, high-density lipoprotein; LDL, low-density lipoprotein; SBP, systolic blood pressure; DBP, diastolic blood pressure; HLP, hyperlipidemia; NLR, neutrophil-to-lymphocyte ratio; HCUCM, The First Affiliated Hospital of Chongqing University of Chinese Medicine; CTMT, Chinese traditional medicine treatment.

Bold values are statistically significant.

The median age of PVD patients at HCQMU was 70 (IQR: 60–77) years and their median age at HCUCM was 72 (IQR: 65–79) years. At HCQMU, 39.5% of the PVD patients (n = 1589) were women and at HCUCM 44.5% (n = 1137) were women. The clinical characteristics at each study center are shown in Supplemental Tables 1 and 2.

For UK biobank validation, our study included 44442 participants to analyze the relationship between the diagnosis of hypoglycemia and PVD after filtration. Among them 2299 cases are defined as PVD. Since the ICD-10 codes are used for hypoglycemia definition, the number of patients with hypoglycemia (n=175 0.4% for non-PVD group and n=165 7.2% for PVD group) significantly decreased when comparing with our dual-center data, where blood glucose monitoring was used as the definition. This is in accordance with the significant difference between hypoglycemia diagnosis rate and detection rate in our dual-center data (Supplemental Figure 1). Other details about the clinical characteristics of participants with and without PVD from UK biobank are shown in Supplemental Table 3.

Logistic analysis of the association of hypoglycemia and PVD

The results of univariable and multivariable logistic regression analysis of the relationship between hypoglycemia and PVD are shown in (Tables 2 and 3).

Table 2.

Logistic regression analysis of patients with and without PVD in both centers.

Variables	Univariable regression		Multivariable regression
Variables	OR (95% CI)	p	OR (95% CI)	p
Hypoglycemia	1.61 (1.48–1.75)	<0.01	1.80 (1.64–1.97)	<0.01
BMI	0.97 (0.96–0.98)	<0.01	0.97 (0.96–0.98)	<0.01
Age	1.04 (1.04–1.04)	<0.01	1.03 (1.03–1.03)	<0.01
Sex	0.85 (0.8–0.9)	<0.01	0.82 (0.76–0.89)	<0.01
HbAlc	1.04 (1.02–1.06)	<0.01	1.11 (1.09–1.13)	<0.01
Hypertension	2.94 (2.75–3.15)	<0.01	2.26 (2.09–2.44)	<0.01
Statin	5.14 (4.82–5.48)	<0.01	3.71 (3.47–3.98)	<0.01
HLP	1.17 (1.11–1.24)	<0.01	1.26 (1.17–1.35)	<0.01
Smoke	1.41 (1.32–1.5)	<0.01	1.17 (1.06–1.29)	<0.01
Alcohol use	1.28 (1.2–1.38)	<0.01	1.05 (0.95–1.16)	0.33
Centers	3.25 (3.05–3.47)	<0.01	2.67 (2.48–2.87)	<0.01

HCQMU is regarded as reference for centers; Male is regarded as reference for sex. Abbreviations: BMI, body mass index; HLP, hyperlipidemia; HCQMU, The First Affiliated Hospital of Chongqing Medical University.

Bold values are statistically significant.

Table 3.

Sensitivity analysis of the association between hypoglycemia and PVD after further adjustments in both centers.

Models	OR	95% CI	p
Model 1	1.80	(1.64, 1.98)	<0.01
Model 2	1.74	(1.59, 1.91)	<0.01
Model 3	1.76	(1.60, 1.93)	<0.01
Model 4	1.64	(1.49, 1.81)	<0.01
Model 5	1.76	(1.61, 1.93)	<0.01

Model 1: Adjusted for age, hemoglobin A1c, study center, body mass index, sex, statin use, smoking, and alcohol use.

Model 2: Model 1 + systolic blood pressure + diastolic blood pressure + hyperlipidemia.

Model 3: Model 1 + total cholesterol + triglycerides + low-density lipoprotein + high-density lipoprotein + hypertension.

Model 4: Model 1 + insulin using + insulin pump use + hypertension + hyperlipidemia.

Model 5: Model 1 + Sulfonylureas using + hypertension + hyperlipidemia.

Bold values are statistically significant.

Both univariable [OR 1.61 (95% CI: 1.48–1.75)] and multivariable analysis [OR 1.80 (95% CI: 1.64–1.97)] found that hypoglycemia and the was significantly associated with PVD (Table 2). The association also remained significant after further adjusting for other potentially related factors such as systolic blood pressure, diastolic blood pressure, low-density lipoprotein, high-density lipoprotein, insulin use, and insulin pump use (Table 3). Logistic regression and the adjusted receiver operating characteristic (ROC) curve for the association of hypoglycemia and PVD are shown in Supplemental Figure 2. The area under the ROC curve was 0.81, with good sensitivity (0.759) and specificity (0.821), indicating a clinically favorable level of differentiation.

For UK biobank validation, the results of univariable and multivariable logistic regression analysis of the relationship between hypoglycemia and PVD are shown in Supplemental Table 4). Both univariable and multivariable analyses found that hypoglycemia was significantly associated with PVD, with an OR of 18.69 (95% CI: 15.04–23.22) in univariable regression and OR 4.58 (95% CI: 3.5–5.99) in multivariable regression (Adjusted for BMI, age, sex, HbAlc, hypertension, HLP, Smoking statue and alcohol use).

Stratified and interaction analysis of the association of hypoglycemia and PVD

The study population was stratified by the values of 11 variables (hypertension, hyperlipidemia, study center, smoking habit, use of traditional medicine, statin use, BMI, age, alcohol use, sex, and IL-6 level. IL-6 values were available from 3900 participants. The interaction effects across each stratum were analyzed by multivariable logistic regression (Figure 3, Supplemental Figure 3).

Figure 3.

Stratified analysis and interaction association of hypoglycemia and PVD (Forest Plot).Adjusted for age, sex, body mass index, hemoglobin A1c, alcohol consumption, smoking, centers, hypertension, hyperlipidemia, and statin use (Factor used for stratification was excluded). Subgroup analysis for each study center was adjusted for all the above factors except the study center.

As shown in Figure 3, multivariable logistic regression consistently indicated a positive association between hypoglycemia and PVD for each stratum. In the stratified analysis, significant interaction associations (p < 0.01) were observed in several subgroups. The association of hypoglycemia with PVD was stronger in individuals with hypertension (OR = 1.98 vs OR = 1.36) or hyperlipidemia (OR = 2.08 vs OR = 1.65) compared to those without those conditions. The effect was also more pronounced in participants older than ≥60 years of age (OR = 2.49 vs OR = 1.67).

IL-6 test results were available from 3,900 patients, and they were stratified into normal and elevated IL-6 level groups (Supplemental Figure 3). The overall analysis also demonstrated a positive relationship of hypoglycemia and PVD. In the stratified analysis, a significant association was observed only in the subgroup with an elevated IL-6 level. Moreover, the interaction analysis confirmed a significant difference between the two subgroups (p = 0.04) (Supplemental Figure 3).

SEM

As shown in Figure 4 and supplemental Table 5, the forward model (Hypoglycemia to PVD) illustrates the relationship between hypoglycemia, NLR, hemoglobin A1c (HbA1c), SBP, and PVD. Hypoglycemia was associated with NLR (coefficient = 0.534), and NLR was further correlated with HbA1c (coefficient = 0.007), while HbA1c showed an additional association with PVD (coefficient = 0.015).Hypoglycemia also related to SBP (coefficient = 1.29), which also contributes to PVD (coefficient = 1.287). Hypoglycemia also influenced PVD without indirect path (coefficient = 0.105). These pathways collectively indicate that hypoglycemia influenced PVD through multiple intermediary variables, including inflammatory markers (NLR), glycemic control (HbA1c), and blood pressure regulation (SBP). The forward model includes cascading associations initiated by hypoglycemia with PVD outcomes. The reverse model (PVD to hypoglycemia) includes the influence of PVD on hypoglycemia and intermediary factors. PVD linked with SBP (coefficient = 2.265) and SBP further associationed the onset of hypoglycemia (coefficient = 0.003). Additionally, PVD was statistically associated with hypoglycemia through an alternative pathway (coefficient = 0.067). This model suggests that PVD was positively influenced by hypoglycemia through both indirect and direct routes.

Figure 4.

Hypothetical structural equation models outlining possible relationships between hypoglycemia and PVD.Paths are adjusted for: age, sex, body mass index, hyperlipidemia, alcohol consumption, smoking, centers, and statin use. Adjusted factors were omitted. Only paths related to the forward or reverse model with p < 0.05 are shown. Abbreviations: E, estimated path coefficient; SBP, systolic blood pressure; PVD: Panvascular disease; HbA1c, hemoglobin A1c; NLR, Neutrophil-to-lymphocyte ratio.

Another models that included patients with IL-6 (n = 3900) and UA (n = 13,052) results also indicate the presence of a correlation structure between hypoglycemia and PVD in the study participants (Supplemental Figure 4a, b) and reinforce the significant roles that metabolic factors and inflammation factors played in both the forward and the reverse association. As shown in Supplemental Figure 4a, the correlation structure between hypoglycemia and PVD was mediated by NLR, HbA1c, UA, and SBP. In the forward model, hypoglycemia had a direct impact on NLR (E = 0.700), which subsequently affected HbA1c (E = 0.013) and UA (E = 0.012). HbA1c contributed to PVD (E = 0.016) and UA directly impacted PVD (E = 0.093). In the reverse model, PVD influenced hypoglycemia by multiple feedback mechanisms. PVD directly impacted UA (E = 0.123), hypoglycemia (E = 0.670), and SBP (E = 4.16). UA (E = 0.067) and HbA1c (E = 0.0003), in turn, directly affected hypoglycemia. Hypoglycemia increased IL-6 (E = 0.067) (Supplemental Figure 4b), which subsequently impacted HbA1c (E = 0.234) and contributed to the development of PVD (E = 0.013). IL-6 also had a direct effect that nearly reached significance (E = 0.023, p = 0.07) on PVD.

Taken together, the forward and reverse models indicate the presence of a bidirectional and interconnected relationship between hypoglycemia and PVD that is mediated by abnormal inflammation, and metabolic abnormalities (glycemic control, uric acid, and blood pressure). The findings are consistent with the existence of a potential mutual association in which hypoglycemia and PVD mutually aggravate each other through complex physiological pathways.

Discussion

In T2DM patients, PVD is more challenging to treat and manage than lesions of single vascular bed. Previous studies show that it also has a poorer prognosis and requires greater clinical attention.^11,14–16 There are metabolic, circulatory, and inflammatory pathways that are common to both PVD and hypoglycemia, suggesting that they may be linked by a bidirectional potential mutual association^{1,5–9,15,22,23} but that relationship is often overlooked in clinical practice.

Our study shows that there was a two-way relationship between hypoglycemia and PVD in these diabetes patients. Using logistic analysis and SEM, our study demonstrated the potential potential mutual association in which these two diseases exacerbate each other through complex physiological pathways (see Tables 2 and 3; Figures 1 and 4; Supplemental Tables 4 and 5; Supplemental Figure 4). Stratified and interaction analysis found that the association of hypoglycemia and PVD was stronger in certain subgroups, particularly those with worse metabolic and inflammatory states (Figure 3, Supplemental Figure 3). In addition, our study also revealed a big gap between the detection rate and diagnosis rate of hypoglycemia, reflecting that hypoglycemia is still underappreciated in clinical practice (Supplemental Figure 1).

Some studies have investigated the link between hyperglycemia and vascular disease.^24–28 About the relationship between hypoglycemia and vascular disease, some studies have explored the inter-relationships of intensive glycemic control, hypoglycemia, and isolated vascular complications that were not comprehensively considered.²⁹ Using data from the ADVANCE trial, Zoungas et al. found that severe hypoglycemia was significantly associated with an increased risk of major macrovascular events (HR, 2.88), major microvascular events (HR, 1.81), cardiovascular death (HR, 2.68), and all-cause mortality (HR, 2.69).³ In the VA Diabetes Trial (VADT), hypoglycemia increased the risk of cardiovascular events to twice that of the control group (HR = 2.00), and also significantly increased the risk of microvascular complications (HR = 1.76). Patients with two or more episodes of hypoglycemia had a higher risk of vascular events compared with those who had one episode (HR = 1.53).³⁰ William et al. reported that intensive glucose control significantly increased the incidence of hypoglycemia and did not significantly reduce vascular events or mortality. The study showed that the risk of hypoglycemia risk might decrease the benefit offered by intensive blood glucose control.³¹ The ACCORD trial also found that intensive therapy group showed no significant reduction in cardiovascular event risk and significantly increased the mortality rate.² Another analysis of VADT trial data found that serious hypoglycemia was associated with progression of coronary artery calcium in the standard therapy group.³² The overall data from previous studies fail to give a comprehensive view of vascular disease, and there is currently no studies of the influence of hypoglycemia on PVD, let alone exploring the reverse pathway are ongoing.

In our study, structural equation modeling (SEM) was used to explore potential inflammatory and metabolic mediators of this relationship, which may help to guide the future research of the underlying mechanisms of the potential mutual association (Figure 4, Supplemental Figure 3). In the forward model, inflammation is the key point of the pathway. Our model included IL-6 and NLR, two inflammatory factors that are strongly associated with diabetic vascular disease and are inter-related with each other.^{19,20,33–35} Following stratification, the association of hypoglycemia and PVD was stronger in individuals with a higher IL-6 level (Supplemental Figure 3). The influence of hypoglycemia on IL-6 is consistent with previous studies.^5,6 There are few published data on the relationship of hypoglycemia and the NLR, but a relevant study reported that the NLR was significantly decreased by ziltivekimab, an antibody targeting the IL-6 ligand.³³ Our study is the first to document the positive association between hypoglycemia and the NLR, which might of use as an inflammation marker in DM patients.

In the reverse model, IL-6 was also a mediator for the effect of PVD on hypoglycemia. The occurrence of PVD is essentially an uncontrolled state of high inflammation. Inflammatory stress leads to worse metabolic disorders, worse blood glucose control, and a greater risk of hypoglycemia.^{11,15,16,36–39} Consistent with previous studies, inflammatory stress had metabolic consequences in our study (elevated HbA1c and UA), ultimately resulting in poorer blood glucose control and hypoglycemia.^36,37 UA is known to be a downstream result of inflammation, to participate in inflammatory processes, and to be a key factor in atherosclerosis.^21,40,41 UA participated in the reverse model, possibly because of the relation of PVD with inflammatory status and PVD-related renal dysfunction that leads to the elevation of UA and a worse metabolic situation.^40,42

SEM suggested the involvement of SBP in the relationship of hypoglycemia with PVD and the stratification analysis also found that the involvement was greater in patients with hypertension. During hypoglycemia, the sympathetic nervous system is overactivated, and the stress leads to increased secretion of catecholamines, resulting in acute blood pressure spikes.^8,43 Our previous study reported that hypoglycemia damaged endothelial function, especially endothelium-dependent vasodilation and finally lead to blood pressure elevation.⁴⁴ High blood pressure, both as a consequence and a cause of vascular complications, exacerbated PVD while increasing the stress response, further stimulating sympathetic activation and the secretion of catecholamines.^45,46 Elevated blood pressure may also increase the risk of hypoglycemia in diabetes patients in several ways, such as the side effects of antihypertensive drugs, the reduction of renal gluconeogenesis and the delayed clearance of insulin and oral hypoglycemic agents because of vascular-related renal dysfunction, and the decrease of counter-regulatory responses and pancreatic insulin secretion due to hypertension-associated sympathetic dysfunction.^45–48

Our study is the first to explore the relationship between hypoglycemia and PVD using a dual-center large sample and to validate the results in a larger multi-ethnic population. Previous studies have always focused on isolated vascular complications. Moreover, this is the first study to describe a bidirectional and interconnected relationship between hypoglycemia and PVD. In clinical practice the management of PVD in patients with T2DM involves multiple departments.^11,14,16 However, due to the separation of these disciplines, specialists often focus on local lesions and neglect the patient’s overall systemic vascular condition. Given that the vascular system functions as an integrated whole, clinical practice should adopt a holistic approach that addresses multiple vascular beds collectively. Last but not the least, the possible potential mutual association of hypoglycemia and PVD might be a key for the management of diabetes patients. Clinicians might be able to improve both sides by breaking this cycle. Consistent with previous studies, our research also highlights the role of overall metabolic and inflammatory dysregulation in both PVD and hypoglycemia and also suggests that hypoglycemia is not merely a consequence of treatment, but also a result of intrinsic dysfunction of the body.

However, there are some limitations. Firstly, given the inherent limitations of a cross-sectional design, temporal ordering cannot be determined and causal pathways cannot be established in the present study. However, these correlations may contribute to further clinical and preclinical studies and provide insight into clinical interventions. Secondly, limited by the retrospective nature of our data, we only focused on selected metabolic and inflammatory markers. However, many other inflammatory and metabolic factors may act as potential bridges between hypoglycemia and PVD. Therefore, future studies need to explore the roles of a broader range of additional potential factors to provide a more comprehensive understanding of their impact. Thirdly, the mediators and potential effect pathways identified through SEM only reflect correlations among the variables and require further validation in basic research. However, in consistent with previous research, it is also an effective and important approach in clinical research, and these potential inflammatory and metabolic markers may help guide future mechanistic studies.¹² Fourthly, we used ICD-10 codes to define panvascular disease, which may introduce some bias. However, using ICD-10 codes to define clinical diseases or outcomes is a commonly used and widely accepted approach in clinical research and it remains the most effective approach for clinical big scale data analysis.^49,50 In future studies, methods such as whole-body vascular imaging should be applied to more accurately diagnose panvascular disease. Lastly, hypoglycemia in this study was defined based on blood glucose measurements obtained during hospitalization, this method may underestimate the true incidence and can not reflect the burden of recurrent or severe hypoglycemia compared to continuous glucose monitoring (CGM). However, in-hospital glucose assessments conducted by healthcare professionals are accurate enough to reflect true hypoglycemic events, which is also suitable and economic

For large scale data analysis. Furthermore, previous studies also showed the limitations of CGM, including issues related to data completeness and validity. As a result, self-monitoring of blood glucose (SMBG) continues to serve as an important standard in the diagnosis of hypoglycemia.¹⁷

Conclusion

Our study firstly described the association of hypoglycemia with PVD, and found the mutually reinforcing correlation structure with a large population and validated it with UK biobank date. Within this potential mutual association, inflammation and metabolism biomarkers act as mediators, potentially serving as focal points for future research. Our study also highlights the insufficient attention clinicians pay to hypoglycemia and PVD, as they often focus only on localized vascular diseases rather than considering the vasculature as a whole and overlooked hypoglycemia. In future clinical practice, greater emphasis should be placed on the relationship of hypoglycemia with PVD, exploring potential approaches to relieve both conditions by breaking this potential mutual association and addressing the mediating roles of inflammation and metabolic dysfunction.

Supplemental material

Supplemental material - The association between hypoglycemia and panvascular disease in type 2 diabetes patients: The correlation structure mediated by inflammatory and metabolic abnormalities

Supplemental material for The association between hypoglycemia and panvascular disease in type 2 diabetes patients: The correlation structure mediated by inflammatory and metabolic abnormalities by Rui Lan, Xunjia Li, Zhulu Chen, Yanwei Li, Bryan Richard Sasmita, Xiankang Hu, Zhixin Xu, Deyu Zuo, Zuo Zhong, An He in Diabetes & Vascular Disease Research.

Footnotes

Acknowledgments

This research has been conducted using the UK Biobank Resource under Application ID 387995.

ORCID iD

An He

Ethical considerations

This study was carried out in accordance with the Declaration of Helsinki, approved by the Ethics Committee of the Chongqing Medical University(ethics number: 2024-034-01) and the Ethics Committee of the The First Affiliated Hospital of Chongqing University of Chinese Medicine (ethic number: 2025-KY-KS-1).

Consent to participate

All the participants provided written permission to use their data for medical research use.

Author contributions

R.L & X.L. & Z.C. led this study, conducted data collection and processing, wrote the manuscript, conducted statistical analysis, and drew tables, figures. Z.C. was the only one who handled all processes related to the UK Biobank data and drew the graphical abstract. Y.L. & T.R. assist data collecting and manuscript submission. Z.X. & B.R.S. & X.H. provided the consultation and revised the manuscript. A.H. & Z.Z. & D.Z. conceived and designed this study and revised the manuscript and provided consultations. All authors have read and agreed to the published version of the manuscript.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research is supported by The National Natural Science Foundation of China [No. 82300477]; The Natural Science Foundation of Chongqing [No. CSTB2024NSCQ-MSX0320]; China Postdoctoral Science Foundation General Funding Program [No. 2024MD754010]; Chongqing Medical Scientific Research project [Joint project of Chongqing Health Commission and Science and Technology Bureau] [No. 2023ZDXM011] and CQMU Program for Youth Innovation in Future Medicine [No. W0188]; The National Natural Science Foundation of China [No. 82305006]; The Chongqing Natural Science Foundation Innovation and Development Joint Fund [Chongqing Education Commission No. CSTB2024NSCQ-LZX0079]; The Chongqing Medical Young Talents Program [No. YXQN202415]; Chongqing Natural Science Foundation [General Project] [No. CSTB2022NSCQ-MSX0081]; Chongqing Postdoctoral Innovative Talent Support Program [CQBX202213].

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The study’s datasets from both centers are not public due to privacy and ethical reasons but can be requested from the corresponding author. The UK biobank data collected for the current study can not be shared without UK Biobank’s explicit written approval.*

Supplemental material

Supplemental material for this article is available online

Appendix

References

Zinman

Marso

Christiansen

, et al. Hypoglycemia, cardiovascular outcomes, and death: the LEADER experience. Diabetes Care 2018; 41(8): 1783–1791.

Action to Control Cardiovascular Risk in Diabetes Study Group . Effects of intensive glucose lowering in type 2 diabetes. N Engl J Med 2008; 358(24): 2545–2559.

Zoungas

Patel

Chalmers

, et al. Severe hypoglycemia and risks of vascular events and death. N Engl J Med 2010; 363(15): 1410–1418.

Seaquist

Anderson

Childs

, et al. Hypoglycemia and diabetes: a report of a workgroup of the American diabetes association and the endocrine society. J Clin Endocrinol Metab 2013; 98(5): 1845–1859.

Hunt

Taber

, et al. Inflammatory biomarkers, glycemic variability, hypoglycemia, and renal transplant outcomes: results of a randomized controlled trial. Transplantation 2014; 98(6): 632–639.

Verhulst

CEM

van Heck

JIP

Fabricius

, et al. Sustained proinflammatory effects of hypoglycemia in people with type 2 diabetes and in people without diabetes. Diabetes 2022; 71(12): 2716–2727.

Papachristoforou

Lambadiari

Maratou

, et al. Association of glycemic indices (hyperglycemia, glucose variability, and hypoglycemia) with oxidative stress and diabetic complications. J Diabetes Res 2020; 2020: 7489795.

International Hypoglycaemia Study

. Hypoglycaemia, cardiovascular disease, and mortality in diabetes: epidemiology, pathogenesis, and management. Lancet Diabetes Endocrinol 2019; 7(5): 385–396.

Lee

Warren

Lee

, et al. The association of severe hypoglycemia with incident cardiovascular events and mortality in adults with type 2 diabetes. Diabetes Care 2018; 41(1): 104–111.

10.

Luo

, et al. Recurrent hypoglycemia impaired vascular function in advanced T2DM rats by inducing pyroptosis. Oxid Med Cell Longev 2022; 2022: 7812407.

11.

Wong

Dai

. Pan-vascular disease: what we have done in the past and what we can do in the future? Cardiology Plus 2024; 9(1): 1–5.

12.

Fisher

Taylor

West

, et al. Bidirectional relationship between chronic kidney and periodontal disease: a study using structural equation modeling. Kidney Int 2011; 79(3): 347–3455.

13.

Zhang

Jia

, et al. Sediments alleviate the inhibition effects of antibiotics on denitrification: functional gene, microbial community, and antibiotic resistance gene analysis. Sci Total Environ 2022; 804: 150092.

14.

Chu

Feng

, et al. Chinese expert consensus on the risk assessment and management of panvascular disease inpatients with type 2 diabetes mellitus (2022 edition). Cardiology Plus 2022; 7(4): 162–177.

15.

Liu

, et al. Diabetic vascular diseases: molecular mechanisms and therapeutic strategies. Signal Transduct Target Ther 2023; 8(1): 152.

16.

Zhou

Zhao

, et al. Panvascular medicine: an emerging discipline focusing on atherosclerotic diseases. Eur Heart J 2022; 43(43): 4528–4531.

17.

Bue-Valleskey

Kazda

, et al. Once-weekly basal insulin Fc demonstrated similar glycemic control to once-daily insulin degludec in insulin-naive patients with type 2 diabetes: a phase 2 randomized control trial. Diabetes Care 2023; 46(5): 1060–1067.

18.

Zhang

Ren

, et al. Unveiling the pathogenesis and therapeutic approaches for diabetic nephropathy: insights from panvascular diseases. Front Endocrinol 2024; 15: 1368481.

19.

Zhang

Wei

Wang

, et al. The neutrophil-to-lymphocyte ratio is associated with all-cause and cardiovascular mortality among individuals with hypertension. Cardiovasc Diabetol 2024; 23(1): 117.

20.

Welty

Hariri

Asbeutah

, et al. Regression of coronary fatty plaque and risk of cardiac events according to blood pressure status: data from a randomized trial of eicosapentaenoic acid and docosahexaenoic acid in patients with coronary artery disease. J Am Heart Assoc 2023; 12(18): e030071.

21.

Kimura

Tsukui

Kono

. Uric acid in inflammation and the pathogenesis of atherosclerosis. Int J Mol Sci 2021; 22(22): 12394.

22.

Zhang

Qin

Zhang

, et al. NDRG1 signaling is essential for endothelial inflammation and vascular remodeling. Circ Res 2023; 132(3): 306–319.

23.

Chen

Zhang

Tang

, et al. Macrophages in cardiovascular diseases: molecular mechanisms and therapeutic targets. Signal Transduct Target Ther 2024; 9(1): 130.

24.

ACCORD Study Group ACCORD Eye Study Group Chew

Ambrosius

Davis

, et al. Effects of medical therapies on retinopathy progression in type 2 diabetes. N Engl J Med 2010; 363(3): 233–244.

25.

Giacco

Brownlee

. Oxidative stress and diabetic complications. Circ Res 2010; 107(9): 1058–1070.

26.

Zhu

Yao

, et al. Hyperglycemia-regulated tRNA-derived fragment tRF-3001a propels neurovascular dysfunction in diabetic mice. Cell Rep Med 2023; 4(10): 101209.

27.

Turnbull

Abraira

Anderson

, et al. Intensive glucose control and macrovascular outcomes in type 2 diabetes. Diabetologia 2009; 52(11): 2288–2298.

28.

Johnston

Bruno

Pauls

, et al. Intensive vs standard treatment of hyperglycemia and functional outcome in patients with acute ischemic stroke: the SHINE randomized clinical trial. JAMA 2019; 322(4): 326–335.

29.

Yang

Fang

, et al. The association between intensive glycemic control and vascular complications in type 2 diabetes mellitus: a meta-analysis. Nutr Metab Cardiovasc Dis 2009; 19(9): 596–603.

30.

Zhao

Campbell

Fonseca

, et al. Impact of hypoglycemia associated with antihyperglycemic medications on vascular risks in veterans with type 2 diabetes. Diabetes Care 2012; 35(5): 1126–1132.

31.

Duckworth

Abraira

Moritz

, et al. Glucose control and vascular complications in veterans with type 2 diabetes. N Engl J Med 2009; 360(2): 129–139.

32.

Saremi

Bahn

Reaven

, et al. A link between hypoglycemia and progression of atherosclerosis in the Veterans Affairs Diabetes Trial (VADT). Diabetes Care 2016; 39(3): 448–454.

33.

Adamstein

Cornel

Davidson

, et al. Association of interleukin 6 inhibition with ziltivekimab and the neutrophil-lymphocyte ratio: a secondary analysis of the RESCUE clinical trial. JAMA Cardiol 2023; 8(2): 177–181.

34.

Buonacera

Stancanelli

Colaci

, et al. Neutrophil to lymphocyte ratio: an emerging marker of the relationships between the immune system and diseases. Int J Mol Sci 2022; 23(7): 3636.

35.

Hunter

Jones

. IL-6 as a keystone cytokine in health and disease. Nat Immunol 2015; 16(5): 448–457.

36.

Hotamisligil

. Inflammation and metabolic disorders. Nature 2006; 444(7121): 860–867.

37.

Deretic

. Autophagy in inflammation, infection, and immunometabolism. Immunity 2021; 54(3): 437–453.

38.

Ilyas

Little

, et al. Endothelial dysfunction in atherosclerotic cardiovascular diseases and beyond: from mechanism to pharmacotherapies. Pharmacol Rev 2021; 73(3): 924–967.

39.

Plummer

Deane

. Dysglycemia and glucose control during sepsis. Clin Chest Med 2016; 37(2): 309–319.

40.

Jung

Kim

, et al. Uric acid and inflammation in kidney disease. Am J Physiol Renal Physiol 2020; 318(6): F1327–F1340.

41.

Ndrepepa

. Uric acid and cardiovascular disease. Clin Chim Acta 2018; 484: 150–163.

42.

Yanai

Adachi

Hakoshima

, et al. Molecular biological and clinical understanding of the pathophysiology and treatments of hyperuricemia and its association with metabolic syndrome, cardiovascular diseases and chronic kidney disease. Int J Mol Sci 2021; 22(17): 9221.

43.

Haas

Borsook

Adler

, et al. Stress, hypoglycemia, and the autonomic nervous system. Auton Neurosci 2022; 240: 102983.

44.

Guo

, et al. Hypoglycaemia aggravates impaired endothelial-dependent vasodilation in diabetes by suppressing endothelial nitric oxide synthase activity and stimulating inducible nitric oxide synthase expression. Microvasc Res 2023; 146: 104468.

45.

McEvoy

McCarthy

Bruno

, et al. 2024 ESC guidelines for the management of elevated blood pressure and hypertension. Eur Heart J 2024; 45(38): 3912–4018.

46.

Díaz-Morales

Baranda-Alonso

Martínez-Salgado

, et al. Renal sympathetic activity: a key modulator of pressure natriuresis in hypertension. Biochem Pharmacol 2023; 208: 115386.

47.

Lee

Wei

, et al. Hypertension in CKD: core curriculum 2019. Am J Kidney Dis 2019; 74(1): 120–131.

48.

Maines

Urru

SAM

Leonardi

, et al. Drug-induced hyperinsulinemic hypoglycemia: an update on pathophysiology and treatment. Rev Endocr Metab Disord 2023; 24(6): 1031–1044.

49.

Vanent

Leasure

Acosta

, et al. Association of chronic kidney disease with risk of intracerebral hemorrhage. JAMA Neurol 2022; 79(9): 911–918.

50.

Thompson

Tong

Shen

, et al. Longitudinal associations between air pollution and incident dementia as mediated by MRI-measured brain volumes in the UK biobank. Environ Int 2025; 195: 109219.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.37 MB