Trajectories of atherosclerotic cardiovascular disease risk scores as a predictor for incident chronic kidney disease

Atherosclerosis is an important contributor to cardiovascular disease (CVD), including myocardial infarction, ischemic stroke, and peripheral vascular disease [1, 2]. Atherosclerotic cardiovascular disease (ASCVD) risk scores have been developed to estimate an individual's risk of CVD [3, 4]. Among them, the revised ASCVD pooled cohort equations, in particular, provide race- and sex-specific 10-year ASCVD risk estimates, highlighting the variability in risk factor prevalence, associated risks, and underlying ASCVD event rates among different groups [3].

There is an inter-relationship between atherosclerosis and renal function [5, 6]. The pathological process of atherosclerosis can stimulate intrarenal atherogenesis through the mediation of oxidative stress and inflammation, subsequently leading to a decline in renal function [5]. Concurrently, a decrease in renal function, especially as seen in chronic kidney disease (CKD), can exacerbate the production of uremic toxins, free radicals, and pro-inflammatory cytokines [6]. Previous research has also suggested that 10-year ASCVD risk scores have predictive power beyond those of conventional cardiovascular risk predictors in determining cardiovascular events in patients with CKD [7‐11]. However, ASCVD risk score changes over time. If temporal trends in ASCVD risk scores can serve as significant predictors of CKD, they could provide additional information for clinicians to identify individuals at risk of CKD, in addition to those at risk of CVD, allowing for the application of cost-effective early interventions. Therefore, this study aimed to investigate the association between the trajectories of ASCVD risk scores and incident CKD using a large community-based prospective Korean cohort.

To the best of our knowledge, this is the first study to investigate the association between ASCVD risk trajectory and incident CKD. Previous studies have concentrated mainly on evaluating ASCVD risk in individuals with pre-existing CKD [23‐25]. However, the identification of individuals at a higher risk of CKD through early detection and implementation of early interventions can be a potentially cost-effective approach to reducing the burden of CKD. In the current study, the ASCVD risk trajectory group had a 2.13 times higher risk of incident CVD than the stable group, regardless of the ASCVD risk at baseline. Moreover, the increasing group consistently showed a higher prevalence of proteinuria than did the stable group during the event accrual period. Mean eGFR was consistently lower in the increasing than in the stable ASCVD risk trajectory group.

Ren et al. [8] found that individuals with higher ASCVD risk scores showed lower eGFR and more severe renal interstitial inflammation among 218 patients with type 2 DM and biopsy-proven CKD and suggested that an ASCVD risk of 14.1% was an independent indicator of renal dysfunction. At the end of the exposure period of the current study, the median ASCVD risk score in the increasing ASCVD risk trajectory group was 12.8. As we investigated the risk of development of CKD in the future using participants without CKD, the ASCVD risk score for predicting renal dysfunction may be lower than that suggested previously.

The risk of incident CKD was associated with ASCVD risk trajectories in the subgroups of patients in their 50s, those without DM, and those without HTN. Particularly, the interaction for the HTN subgroup was significant, suggesting that the presence of HTN should be prioritized when evaluating the risk of CKD using ASCVD risk trajectories. This finding highlights the critical need to incorporate HTN status into the assessment models, emphasizing its substantial impact on the progression and risk evaluation of CKD within individualized patient care strategies. The incidence of CVD is approximately 40% in the age group of 40–59 years and 75% in the age group of 60–79 years [26]. In a study that tracked middle-aged men over a period of 40 years, the influence of traditional CVD risk factors declined with age and only low-density lipoprotein cholesterol consistently affected CVD risk, while BMI and FPG had an impact on heart failure, regardless of age [27]. Given that age in itself is a significant risk factor for both CVD and CKD, the lack of a significant correlation between the increase in ASCVD risk score in individuals in their 60 s and the risk of incident CKD in our study could suggest that age may have played a vital role in the development of CKD. In contrast, the incidence rates of CKD per 100,000 population years were 132 for mild CKD, 33 for moderate CKD, and 15 for severe CKD in individuals aged 20 to 64 years [28]. Participants from the current study aged 40–49 years may not have been able to demonstrate the impact of changes in the ASCVD risk score on the incidence of CKD due to low CKD rates even after a period of 16 years, given that they remained below the age of 65 at completion of the study. In contrast, we hypothesized that, in individuals aged 50–59 years, we were able to establish a significant correlation between alterations in ASCVD risk scores and the incidence of CKD. This is likely related to their reaching an increased risk of CKD during the substantial 16-year follow-up period, whereas the effects of CVD risk factors did not diminish with age. In patients with DM, those individuals with DM at baseline could have already developed CKD to a sufficient extent during the exposure period, given that it takes approximately 7–10 years for CKD to develop [29]. The rapid decline in renal function was observed particularly in patients with DM with high SBP [30]. Similarly, in patients with HTN, an analysis using data from a total of 43,305 patients with HTN from the United States revealed that time-varying SBP is a factor for a rapid decline in eGFR, with every 10 mmHg increase in SBP leading to a 0.2 mL/min/1.73 m² decline in eGFR [31]. As a consequence, individuals exhibiting an upward trajectory in ASCVD risk scores may have been excluded due to the premature development of CKD. Furthermore, given that the formula for calculating ASCVD risk scores incorporates both DM and SBP, this inclusion could potentially diminish the observed association between variations in ASCVD risk scores and the onset of CKD during the period of event accumulation in patients diagnosed with either DM or HTN. The absence of a significant association between ASCVD risk score changes and CKD onset in the DM and HTN subgroups in this study could be due to this factor.

We suggest several explanations for these findings, as well as providing the exact underlying mechanism for the significant relationship between ASCVD risk trajectory groups and incident CKD. First, it is possible that individuals in the group with increased ASCVD risk experience comorbidities such as HTN, DM, and dyslipidemia [32], leading to an increased risk of CKD. At the molecular level, a recent study suggested that microRNAs (miRNAs) mediate homeostasis of the vascular wall and are involved in vasculature pathologies [33]. miR-155 and miR-223 play a role in the imbalance of calcium and phosphate in the vessels, which leads to the loss of these minerals from the bones and their accumulation in the vascular wall, contributing to vascular calcification. Additionally, miR-21 is upregulated in endothelial cells from atheroma plaques and progenitor cells from patients with coronary artery disease and is associated with the physiological proliferation of vascular smooth muscle cells and kidney fibrosis [30]. Second, an increase in ASCVD risk could reflect a more active atherosclerotic disease process, thus leading to higher susceptibility to renal damage and subsequent CKD [5, 34]. Cigarette smoking can exacerbate endothelial dysfunction and facilitate the infiltration of inflammatory cells into the vascular wall, thus promoting the formation of atherosclerotic plaques. In the current study, a higher proportion of daily smokers and elevated CRP levels were seen in the group with increasing ASCVD risk trajectory than that in the stable group. This could contribute to the observed association between an increased ASCVD risk trajectory and CKD incidence. Additionally, other lifestyle factors such as physical inactivity and alcohol consumption may contribute to the progression of ASCVD, which may lead to renal damage over time.

This study had several limitations. First, the exclusion of individuals with new-onset CKD during the exposure period may have limited the ability of our trajectory modeling to fully reflect the risk of CKD onset. Second, the risk of CKD may have been underestimated in the increasing ASCVD risk trajectory group. Additionally, a more accurate definition of CKD could have been achieved if not only the GFR category but also the cause and albuminuria category had been utilized. However, due to the limitations of the cohort, including a lack of information on the cause of CKD and the potential inaccuracy of proteinuria measurements in individuals taking antihypertensive medications, our study defined CKD based solely on the GFR category. Third, our results may not be applicable to other ethnicities because the KoGES consists only of Koreans. Finally, we focused solely on changes in ASCVD risk scores over time and did not account for the dynamic effects of other important variables such as BMI, blood pressure, plasma glucose level, lipid profiles, and lifestyle factors. Further studies incorporating these factors and their time-dependent effects will provide valuable insights into the development and progression of CKD.

In conclusion, our study provides evidence that an increasing trend in ASCVD risk scores serves as an independent predictor of adverse renal outcomes such as new-onset CKD and the presence of proteinuria, in patients without DM or HTN. Our findings underscore the importance of continuous monitoring and effective control of ASCVD risk factors, not only for cardiovascular health but also for the prevention and management of CKD. Further research is needed to validate these findings and elucidate the precise mechanisms underlying this association.