Study cohort

Being a prospective cohort study, health examinations for active and retired Kailuan Group personnel were conducted every two years from June 2006 to October 2007 in 11 hospitals, including Kailuan General Hospital and its affiliated hospitals. Follow-up assessments included incident HF and mortality. In the 6th health examination in 2016, cystatin C was investigated in T2D patients. We selected T2D patients who participated in this health examination and underwent cystatin C investigations as our study subjects. The inclusion criteria were: (1) Those who participated in the 2016 annual health examination and met the diagnostic criteria for T2D; (2) Patients with the availability of primary research data, including cystatin C and serum creatinine, and (3) Those willing to participate and provide informed consent. The exclusion criteria were: (1) Patients with a history of HF before the health examination and (2) Those having valvular and congenital heart diseases, respectively.

Collection of general clinical data and laboratory investigations

Patient data like age, gender, personal history, disease history, and medication usage were obtained through face-to-face interviews. We measured height, weight, blood pressure, heart rate, and relevant biochemical indicators by following previously published methods [16]. Smoking was defined as averaging at least one cigarette per day > 1 year or having quit smoking < 1 year ago. Additionally, body mass index (BMI) was calculated as weight / height2 (kg/m2).

Calculation and grouping of eGFRcr and eGFRcys

We used the 2012 CKD Epidemiology Collaboration (CKD-EPI) cystatin C equation and 2021 race-free CKD-EPI equations to calculate eGFRcys and eGFRcr, respectively [10, 17], and eGFRdiff = eGFRcys- eGFRcr.

The subjects were divided into 3 groups according to eGFRdiff level: negative eGFRdiff group: lower than − 15 mL/min/1.73 m2, with eGFRcys lower than eGFRcr; midrange eGFRdiff group: -15 to 15 mL/min/1.73 m2, with eGFRcys similar to eGFRcr; positive eGFRdiff group: 15 mL/min/1.73 m2 or greater, with eGFRcys higher than eGFRcr.

Diagnostic criteria.

T2D: The American Diabetes Association (ADA) Criteria for Diagnosis of Diabetes (2010) was referred [18].

1) History of T2D;

Or 2) Fasting blood glucose (FBG) ≥ 7.0 mmol/L;

Or 3) Two-hour blood glucose of ≥ 11.1 mmol/L in random plasma glucose test or oral glucose tolerance test;

Or 4) Hemoglobin A1c (HbA1c) ≥ 6.5%(47.5 mmol/mol).

HF: Chinese Guidelines for the Diagnosis and Treatment of Chronic Heart Failure (2018) was referred [19].

1)

Symptoms and signs of HF, manifested as shortness of breath, fatigue, palpitations, fluid retention, as well as New York Heart Association (NYHA) heart function grade II and above;

2)

Modified Simpson’s method: the left ventricular ejection fraction < 50% measured by echocardiography;

3)

Plasma N-terminal pro-B-type natriuretic peptide ≥ 125 ng/L.

The diagnosis must meet conditions (1) as well as at least one of conditions (2) and (3).

Follow-up and endpoint events

After the completion of the 6th health examination, that is, the starting point of follow-up, trained medical staff reviewed the inpatient diagnosis and recorded the end-point events of the participants in the Affiliated Hospitals of Kailuan Group and the Designated Hospitals for Medical and Health Insurance of China every year. The end-point events ware defined as HF during the follow-up. The time of the first event was considered as the end-point for those with > 2 events, and the final follow-up date for those without HF was December 31, 2020. All diagnoses were confirmed by professional physicians according to the inpatient medical records.

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 HF events in each group and the overall population, and a log-rank test was adopted to compare the difference in the incidence of HF.

eGFRcys and eGFRcr were categorized into the following four groups (ml/min/1.73m2 ): ≥90, 60–89, 45–59, ≤ 45.The eGFRdiff was assessed as a categorical and continuous variable. The effect of different eGFR groups and each 15 increases in eGFR on new-onset HF was studied using a multivariate Cox stepwise regression model. Model 1 unadjusted. Model 2 was adjusted for age and sex. Model 3 was further adjusted for SBP, BMI, TC, HbA1c, hemoglobin, smoking, anti-diabetic treatment, antihypertensive treatment, MI, and atrial fibrillation.

In addition, based on Model 1(age, sex), Model 2 (ARIC-sans-BNP model: age, sex, HR, BMI, SBP, HbA1c, hypertension and MI), C-Statistic, net reclassification index (NRI), and integrated discrimination improvement (IDI) were used to assess the ability of different eGFR to improve HF prediction models, respectively.

In order to avoid the influence of MI and hypertension on HF, sensitivity analysis was performed after excluding the above population.

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.