Introduction
China is a significant contributor to the burden of cardia-cerebrovascular diseases (CVD), with approximately 330 million individuals currently affected by CVD, including around 13 million stroke cases and 11.39 million cases of coronary heart disease [
1]. The country is facing the dual challenges of an aging population and the persistent prevalence of metabolic risk factors, leading to a continuous increase in CVD prevalence and the highest mortality rate. As certain risk factors may be asymptomatic, patients may already have developed vasculopathy by the time these risk factors are detected, resulting in more severe events such as myocardial infarction or stroke. Diabetes is a major risk factor for CVD, and individuals with diabetes are considered a high-risk population for CVD. Currently, the prevalence of diabetes in China is 12.8% (according to ADA criteria), with a total of 129.8 million adults affected [
2]. Therefore, the key focus of CVD prevention is to identify patients at high risk and implement appropriate interventions.
Diabetic microangiopathy may be one of the mechanisms of high risk of CVD. Albuminuria testing, specifically the measurement of urinary albumin-to-creatinine ratio (uACR), has been recommended to assess renal function in patients with type 2 diabetes(T2D) [
3]. Albuminuria, an early-stage marker of blood vessel dysfunction and a sensitive indicator of diabetic microangiopathy, is increasingly recognized as a predictive indicator of cardiovascular risk and all-cause mortality in numerous studies [
4]. Most studies have confirmed that elevated uACR is an independent risk factor for increased morbidity and mortality of CVD, whether in the general population [
5‐
7], individuals at low risk of CVD (without diabetes or hypertension) [
8,
9], or individuals with diabetes who have a high risk of CVD (8). Some research even suggests that this conclusion holds true when uACR is below the clinical threshold for albuminuria [
10‐
12]. However, certain other studies have shown no significant difference between uACR and CVD in a cross-sectional analysis [
13,
14].
The existing studies only examined the relationship between uACR and the overall occurrence of CVD events. However, no studies have explored the association between uACR and specific CVD events in the diabetic population. Therefore, in this study, we investigated the relationship between uACR in individuals with T2D and both the total number of CVD events and specific types of CVD events using data from the Kailuan study. Furthermore, we discussed the impact of incorporating uACR into the CVD prediction model.
Methods
Study populations
This prospective cohort study comprised in-service and retired Kailuan employees of the Kailuan Group, who participated in the physical examination conducted every two years in Kailuan General Hospital and the affiliated hospitals from June 2006 to October 2007. The follow-up included an evaluation of myocardial infarction, ischemic and hemorrhagic stroke. As urine albumin and creatinine tests were added during the physical examinations in 2014 (5th) and 2016 (6th), diabetic patients who underwent these tests and participated in the 5th and/or 6th physical examinations were enrolled.
Inclusion criteria: (1) those who participated in the 5th or 6th physical examination; (2) Participants who met the diagnostic criteria for type 2 diabetes; (3) those who had complete urine albumin and creatinine data, and (4) those who agreed for participation and signed informed consent. Exclusion criteria: (1) those with a previous history of CVD (including myocardial infarction, ischemic and hemorrhagic stroke); (2) those who did not agree to participate in this study. This study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of Kailuan General Hospital.
Epidemiological investigations and biochemical and anthropological measurements were detailed in the published literature [
15]. Subjects sat still for 15 min before measuring their blood pressure. A bench- top mercurial phygmomanometer was employed to measure the right brachial pressure. Three consecutive measurements were taken with an interval of 1–2 min between each measurement, and the average of the three measurements was considered. Smoking was defined as an average of ≥ 1 cigarette/day in the last year. Body mass index (BMI) = weight (kg)/height (m)
2. The estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation [
16].
Urine albumin and urine creatinine determination and grouping
After an overnight fast, a single random midstream morning urine sample was collected. All participants’ morning urine samples were centrifuged at 600 g for 5 min and stored at − 80 °C until tested. A urine analyzer was used to measure all of the urine samples (N-600, Dirui, Changchun, China). Jaffe’s kinetic method was used to measure urinary creatinine. Turbidimetry was used to measure urinary albumin (DAKO kit, Denmark).
We looked at uACR as a continuous and categorical variable, with normal (uACR < 3 mg/mmol), microalbuminuria (3–30 mg/mmol), and macroalbuminuria (≥ 30 mg/mmol) categories [
17,
18].
Follow-up and end‐point event
After the completion of uACR determination, that is, the starting point of follow-up, trained medical staff reviewed the inpatient diagnosis and recorded the end‐point events of the observation objects in the Affiliated Hospitals of Kailuan Group and the Designated Hospitals for Medical and Health Insurance of China every year. The end‐point events were defined as CVD during the follow‐up, including myocardial infarction and ischemic and hemorrhagic stroke (please refer to the Standards from World Health Organization for their definitions and diagnostic criteria). Based on the inpatient medical records, professional doctors confirmed all diagnoses. The time and event of the first event were considered as the endpoint for those with ≥ 2 events, and the final follow‐up date for those without events was December 31, 2020.
Statistical analysis
Normally distributed measurement data were expressed as mean + sd. Multiple pairwise-comparison between different groups was conducted using a one-way analysis of variance. The least significant difference (LSD) test and Dunnett’s T3 test were used for evaluating the homogeneity of variance and heterogeneity of variance, respectively. Non-normally distributed data were presented as median and centiles (25th and 75th), while the comparison between the groups was performed using the Kruskal-Wallis rank sum test. Enumeration data were presented as frequency and percentage (n, %), and comparisons between groups were performed by the chi-square test. The Kaplan-Meier method was used to calculate the incidence of CVD events in each group and the overall population, and a log-rank test was adopted to compare the difference in the incidence of CVD.
The uACR was assessed as a categorical and continuous variable. Given a non-normal distribution, uACR was ln-transformed for the continuous model. The effect of different uACR groups and each 1-standard deviation (SD) increase in ln (uACR) on new-onset CVD events was studied using a multivariate Cox stepwise regression model. Model 1 unadjusted. Model 2 was adjusted for age and gender. Model 3 was further adjusted for SBP, FBG, LDL cholesterol, BMI, eGFR, smoking, anti-diabetic treatment and antihypertensive treatment.
In addition, based on Model 2(age, gender), Model 4 (HKU-SG risk score: age, gender, SBP, DBP, HbA1c, LDL cholesterol, BMI, CKD (evaluated by eGFR), atrial fibrillation and smoking), the receiver operating characteristic (ROC) area under the curve (AUC), net reclassification index (NRI), and integrated discrimination improvement (IDI) were used to assess the ability of uACR to improve CVD prediction models, respectively.
A spline function curve was plotted to see if there was a linear correlation between uACR and new-onset CVD events. The multivariable adjusted model include age, gender, SBP, FBG, LDL cholesterol, BMI, eGFR, smoking, anti-diabetic treatment and antihypertensive treatment.
SAS version 9.4 was used for the analysis (SAS Institute, Cary, NC, USA). All statistical analyses were double-tailed, with statistical significance set at P < 0.05.
Discussion
This study indicates that in patients with T2D, an increase in uACR is an independent risk factor for myocardial infarction, ischemic and hemorrhagic stroke, as well as total CVD events. The relationship between uACR and total CVD events remains consistent across different populations, including those with varying genders, BMI, eGFR, and the presence of hypertension. Moreover, there is a dose-response relationship between uACR levels and the incidence risk of CVD events. Furthermore, incorporating uACR into established CVD risk prediction models improves the accuracy of predicting CVD risk.
Previous research have confirmed the relationship between uACR and CVD in the general population and in people with diabetes [
12,
20], but above researches focused on total CVD events as endpoints, yielding inconsistent results. Results from the HOPE study [
12] indicated that for every 0.4 mg/mmol increase in uACR, the risk of CVD increased by 5.9% after adjusting for age and sex in individuals with or without diabetes. The SHS [
10] observed among all partticipants, risk of developing all CVD events increased by 13% for every doubling of uACR within the normal range, but among participants with diabetes, risk of all CVD events increased by 20%. A meta-analysis [
6] confirmed that high uACR was associated with increased risk of ischemic stroke (HR, 1.60; 95% CI: 1.43–1.80), as well as hemorrhagic stroke (HR, 1.76; 95% CI: 1.22–1.45). Whereas subgroup analysis revealed high uACR was unable to predict stroke in patients with T2DM (HR, 2.25; 95% CI: 0.55–9.17). Aguilar [
7] indicated among community-dwelling older adults, uACR was strongly associated with risk of incident stoke of any type and ischemic strok, but not hemorrhagic stroke. However, this study not only confirms the independent association between increased uACR and long-term incident total CVD events in patients with T2D but also examines the relationship between uACR and different types of CVD events. Additionally, our previous research has already demonstrated the link between increased uACR and an elevated long-term risk of heart failure in patients with T2D [
21].
Specifically, this study reveals that in T2D patients with microalbuminuria (uACR levels ranging from 3 to 30 mg/mmol), the incidence risk of developing myocardial infarction, cerebral infarction, and total CVD events increases by 2.86-fold (95% CI: 1.04–2.37), 2.46-fold (95% CI: 1.00–1.54), and 1.30-fold (95% CI: 1.07–1.57), respectively. For patients with macroalbuminuria, the risk escalates by 2.86-fold (95% CI: 1.63–5.00), 2.46-fold (95% CI: 1.83–3.30), and 2.42-fold (95% CI: 2.42–3.15), respectively. Notably, when compared to individuals with normal uACR levels, those with macroalbuminuria have a significantly increased risk of hemorrhagic stroke by 4.69-fold (95% CI: 1.72–12.78), while the risk is not significantly different for individuals with microalbuminuria.
Our findings demonstrate that an increase in uACR is not only an independent risk factor for the development of CVD events in patients with T2D but also shows a dose-response relationship with the risk of CVD. Even when the level of albuminuria below the clinical threshold value of 3 mg/mmol, for the patients with T2D, an increase in uACR is associated with a significant elevation in CVD risk, which is consistent with previous studies [
8,
11,
22]. Additionally, we observed that as uACR increased to a certain level, the growth trend of the risk for each specific CVD event and total CVD events became less steep than before. Similar findings were reported in the ARIC study [
13], where an increased risk of heart failure was observed at a relatively high normal uACR level (approximately 1–3 mg/mol), with a relatively slower growth trend. These studies suggest that for the prevention of long-term CVD events in diabetes patients, a lower uACR level may be preferable.
In recent years, numerous epidemiological surveys and clinical studies on CVD risk factors have indicated that, in addition to traditional factors such as age, coronary heart disease, hypertension, and hyperglycemia, several other factors require further investigation and validation due to their close association with CVD incidence. The PREVEND study [
23] showed that albuminuria, measured in 24-hour urine samples, is associated independently with cardiovascular outcomes (including myocardial infarction and stroke) in the general population and adding albuminuria to a model consisting of Framingham risk factors significantly contributed to identifying individuals at risk of cardiovascular outcomes. Currently, urinary albumin testing(spot urine sample) is already recommended for all patients with T2D to assess for chronic kidney disease. uACR could therefore readily be used as a more formal tool for cardiovascular risk prognostication. Our study confirms that incorporating uACR into established CVD risk prediction models, specifically by integrating uACR into the HKU-SG risk scoring, can enhance the accuracy of predicting the risk of developing CVD. The SAVOR-TIMI53 Trial [
24] found that in patients with T2D, uACR was independently associated with increased risk for cardiovascular outcomes (cardiovascular death, myocardial infarction, or ischemic stroke), and uACR offers incremental prognositic benefit when cardiac biomarker was not considered. These results suggest that uACR can provide a predictive value beyond the traditional risk factors for CVD in patients with T2D, so uACR should be monitored regularly in the early stages of diabetes.
The underlying mechanisms underlying the associations between albuminuria and CVD risk are not well established. However, abnormal albuminuria indicates generalized vascular dysfunction and is related to systemic and coronary atherosclerosis [
25,
26]. In patients with T2D, increased uACR are likely an early signal of microvascular disease and indicate some degree of kidney damage [
27].the widespread vascular disorder may progress to loss of vessel distension and generalized elevation of arterial blood pressure, ultimately predispose patients to the development of micro- and macrovascular disease [
11]. Consistent with previous studies, we also found that the effect of uACR on CVD events was independent of eGFR, these findings may support the hypothesis that albuminuria conferring risk of incident CVD is independent of renal filtration function [
7]. Moreover, common risk factors may underlie the association between albuminuria and CVD events [
28].However, even after full adjustment of conventional cardiovascular risk factors and other potential confounders, albuminuria was still significantly associated with high CVD risk in the present study, suggesting that independent and additional mechanisms may be involved. Also, the association between albuminuria and CVD events is probably explained by a common pathophysiologic process, such as endothelial dysfunction [
29] or chronic, low-grade inflammation [
30]. Further studies are clearly required to expand our understanding in this field.
Prominent strengths of this study are its prospective design, large sample size, multivariable-adjusted analyses and continuous surveillance and careful confirmation of incident CVD events. However, our study also has several limitations. First, uACR was only measured once at baseline in this study; thus, we cannot exclude intrapatient sampling variability. Secondly, the identification of myocardial infarction and stroke cases was based on hospitalization codes, but we did not include clinical data related to specific symptoms of CVD, which may have excluded patients who did not receive hospital treatment and did not provide a more holistic understanding of the study population’s cardiovascular health. Moreover, the duration of follow-up in our study was relatively short, with a median follow-up time of 4.05 years, and therefore, some endpoint events may not have fully occurred. Also, the number of events for hemorrhagic stroke was small, resulting in compromised ability to adjust for covariates and wide Cis and thus needing caution when interpreting relevant results. Finally, because the study participants were mostly male Kailuan Group employees, the extrapolation of results may be limited. However, the results in the male and female populations were both consistent with those in the overall population after gender subgrouping.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.