Study population
The participants were from the Uppsala site (n = 5,036) and Malmö site (n = 6,251) of the Swedish CArdioPulmonary BioImage Study (SCAPIS) cohort, a population-based study of cardiovascular and pulmonary diseases and their risk factors. In total, SCAPIS recruited 30,154 individuals aged 50–64 randomly from geographical areas surrounding 6 university hospitals across Sweden from 2013 to 2018. Each individual donated blood and urine for biobanking, questionnaires were administered and computed tomography imaging was performed. The biological samples were kept frozen at -80 degrees Celsius until analysis. The informed consent was obtained from all subjects and/or their legal guardian. The SCAPIS has been performed in accordance with the Declaration of Helsinki, and was approved by the ethical review board at Umeå University, Sweden (2010-228-31 M), and the ethical boards approved urine samples at Uppsala University and Lund University, Sweden (EPN Uppsala University 2016/387, and 2018/315; Lund University 2016/1031).
Coronary atherosclerosis
The coronary atherosclerotic burden was measured with CCTA and CACS according to Agatston (Somatom Definition Flash, Siemens Medical Solutions, Solna, Sweden). Details regarding cardiac imaging have previously been described15.
The values of CCTA were defined as “No stenosis”, “Non-significant stenosis (< 50%)”, and “Significant stenosis ( > = 50%)”, by measuring the stenosis status of 18 segments of coronary arteries visually by experienced radiologists and cardiologists. Participants who had a stent in an artery or ever underwent CABG (Coronary Artery Bypass Graft) were defined as having significant stenosis > = 50%. If the calcification blooming was difficult to evaluate, it was defined as non-significant stenosis < 50%. Non-contrast enhanced images were applied to measure the total amount of calcifications in each artery and were summed to a total CACS according to international standards16. We divided the sum from CACS into five categories usually used in clinical practice (0, 0.1–100, 101–400, > 400). The detailed methods have been previously reported26.
Assessment of urinary kidney damage biomarkers and covariates
At the first visit, venous blood and spot urine samples were collected after an 8-hour overnight fast. Spot urine samples were aliquoted and frozen at – 80 °C within two hours. Blood samples were immediately analyzed at the university hospital laboratory. Creatinine, glucose, and low-density lipoprotein cholesterol (LDL-C) was determined by direct measurement using Cobas 501 (Roche Diagnostics, Solna, Sweden). The urine sample was sent to the University Hospital of Uppsala for further analysis of the urinary kidney damage biomarkers. Urinary KIM-1 (DY1750B), Osteopontin (DY1433), DKK-3 (DY1118), and EGF (DY236) were analyzed by commercial sandwich ELISAs (R&D Systems, Minneapolis, MN, USA) according to the instructions of the manufacturer. Osteopontin and DKK-3 were analyzed in both cohorts, whereas KIM-1 and EGF were only analyzed in samples from the Uppsala site.
A value of urinary kidney damage biomarkers lower than the lower limit of quantification (LLQ) was imputed by LLQ/√2. The urinary biomarker to U-creatinine ratio was calculated for each biomarker (KIM-1, osteopontin, DKK, EGF, albumin) for adjusting urinary concentration. The CKD-EPI was used for calculating eGFR from serum creatinine, age and sex17.
Other covariates
A validated and standardized questionnaire was used to collect sociodemographic, lifestyle, health, previous comorbidities, medical treatment for underlying diseases, and cardiovascular risk factors information in the SCAPIS population26. Recruitment center was also categorized as binary according to whether the participant was included in Malmö or Uppsala centers. Self-reported smoking status was categorized as never, former, and current smokers. Body mass index (BMI, kg/m2) was estimated by dividing weight (measured in kg) over the square of the height (measured in meters).
Statistical analysis
Continuous baseline variables were presented as means (standard deviations) or median (interquartile range), while categorical variables were expressed as n (%). An ordered logistic regression analyzed the association between atherosclerosis and urinary biomarkers. Atherosclerosis levels were categorized into three groups in CCTA and four groups in CACS. To control for confounding, three models were used: model 1 adjusted for age, sex, country of birth, and individual cohort adjustments; model 2 further adjusted for eGFR and albuminuria; and model 3 adjusted for LDL-C, systolic and diastolic blood pressure, antihypertensive and anti-hyperlipidemia treatment, diabetes mellitus diagnosis, antidiabetic treatment, and smoking status. Sensitivity analyses were conducted using a generalized ordered logistic regression model to assess robustness and account for potential violations of assumptions. Forest plots illustrated the OR of atherosclerosis levels in both the main analysis and sensitivity analyses. Sensitivity analyses were performed for participants with specific criteria, including eGFR > 60 ml/min/1.73 m2, normoalbuminuria, self-reported non-hypertension, non-diabetes, and no known cardiovascular disease. Logistic regression assessed model performance using C-statistics and likelihood ratio test, comparing CCTA levels of ≥ 50% versus < 50% or no stenosis. We conducted further Spearman correlation analysis to explore the relationship between urinary biomarker levels, eGFR, and albuminuria. Statistical significance was set at a two-sided p-value < 0.05. All analyses utilized R (version 4.0.1) and Stata (version 15, College Station, TX, USA).
link