Association Between Cardiovascular Biological Age and Cardiovascular Disease　― A Prospective Cohort Study ―

Abstract

Background: Biological age serves as a common starting point for various age-related diseases and can be associated with a wide range of cardiovascular outcomes. However, associations between cardiovascular biological age (CBA) and various types of cardiovascular disease (CVD) remain unclear.

Methods and Results: Analyzing 262,343 UK Biobank participants, we constructed CBA based on composite biomarkers using the Klemera-Doubal method (denoted as KDM-CBA). We measured KDM-CBA acceleration as the difference between KDM-CBA and chronological age. We then examined the associations between KDM-CBA and 17 CVD types using Cox proportional hazard models. We used restricted cubic spline models to assess potential nonlinear associations of KDM-CBA and KDM-CBA acceleration with different types of CVDs. We observed that KDM-CBA (per 1SD increase) was associated with various CVD types, but with different extent (hypertension: hazard ratio (HR)=2.115, 95% confidence interval (CI): 2.083–2.148; coronary atherosclerosis: HR=1.711, 95% CI: 1.545–1.896). We observed similar results for KDM-CBA acceleration and KDM-CBA. KDM-CBA and KDM-CBA acceleration showed J-type nonlinear associations with nearly all CVD types (cutoff values of ≈55 and −1.7 years for KDM-CBA and KDM-CBA acceleration, respectively).

Conclusions: Our study showed that CBA is associated with increased incidence of CVD, which further validates aging as a common starting point for different CVD types as well as highlighting CBA’s role as an early CVD indicator, providing valuable insights for CVD interventions.

Cardiovascular disease (CVD) has become the leading cause of death worldwide, and aging carries a significant weight in its occurrence and progression.^1–4 Biological aging is a nonlinear change over time that better characterizes aging than chronological age. It is also the underlying process of age-related susceptibility to the development of disease, disability and dysfunction, which further increases mortality rates, commonly portrayed in terms of biological age (BA).^1,2,5,6

The measures of BA include DNA methylation, transcriptomics, proteomics, frailty index and composite biomarkers.^7,8 Composite biomarkers, as sensitive indicators for evaluating health and disease states, are accessible, low-cost and facilitate universal screening of large-sample populations.⁹ However, different organs and systems have different rates of biological aging.¹⁰ Current studies on composite biomarkers have mainly focused on general whole-body BA based on multi-organ systems, which may not accurately capture the associations of various CVDs.

Cardiovascular BA (CBA) assesses the biological aging status of the cardiovascular system. However, research on CBA, constructed using composite biomarkers reflecting the biological aging of the cardiovascular system, is still insufficient. Only 2 studies have measured and investigated CBA, suggesting it to be a superior indicator of CVD-induced death compared with liver and kidney BA.^2,10 However, the range of biomarkers used in constructing CBA is relatively limited and may not provide a comprehensive picture of CBA. Therefore, there is a need to construct a specific CBA using more comprehensive clinical composite biomarkers to provide an early indicator of biological aging of cardiovascular system.

BA serves as a common starting point for various age-related diseases and can be associated with a wide range of cardiovascular outcomes.¹¹ Previous studies have mainly reported on BA, which is associated with the risk of CVD and can capture differences in cardiovascular health,¹² yet few studies have concentrated on CBA. Furthermore, even though a few studies have explored the associations between CBA and cardiovascular outcomes, most have focused on specific or single outcomes, mainly including ischemic heart disease, hypertension or stroke.^10,13,14 In addition, the aging state may also be induced by oxidative stress, metabolic dysfunction, and aging of different cardiac cell types, which can direct the pathophysiology of CVD.^15,16 Different types of CVD (e.g., cardiac arrest, chronic ischemic heart disease, etc.) have different pathogenic mechanisms, which have not been fully considered in previous studies. Given the available evidence, it is reasonable to hypothesize that CBA can serve as a common starting point for CVD types but with heterogeneity across types.

To address the identified research gaps, we conducted research based on UK Biobank, aiming to answer the following questions. Firstly, how is CBA constructed using comprehensive biomarkers to specifically portray the biological aging of cardiovascular function? Secondly, does CBA serve as a common starting point for different types of CVD and does it characterize the heterogeneity of different types of CVD?

Methods

Study Design and Population

The study design is shown in Figure 1. In Part I, we constructed CBA using composite biomarkers (including all blood assay and anthropometric data that can reflect the biological aging of cardiovascular function) using the Klemera-Doubal method (KDM), denoted as KDM-CBA. In Part II, we analyzed the associations between CBA and various types of CVD. We also used restricted cubic spline models to assess potential nonlinear associations of KDM-CBA with different types of CVD.

Figure 1.

Design diagram for this study. In Part I, we constructed the cardiovascular biological age (CBA) using composite biomarkers (including all blood assay and anthropometric data that reflect the biological aging of cardiovascular function) using the Klemera-Doubal method (KDM). In Part II, we analyzed the associations between CBA and various types of CVD using Cox proportional risk models. In addition restricted cubic spline was used to explore potential non-linear associations. CI, confidence interval; HR, hazard ratio.

We based study on the UK Biobank. As a prospective cohort study, the UK Biobank is a rich phenotypic resource that includes lifestyle, biomarkers, and clinical outcomes from >500,000 participants aged 37–73 years, with baseline survey data collected between 2006 and 2010. All participants underwent a series of assessments at baseline, including electronic questionnaires, anthropometric measurements, comprehensive medical examination, and biochemical index tests. All participants signed informed consent before data collection, and enrollees were followed up for disease morbidity and mortality based on UK national death, cancer registries, and hospital admission records.

The study excluded participants who lacked information on: biomarkers for the construction of KDM-CBA (n=69,656), self-reported history of CVD, or a hospital medical record of CVD (n=162,153) at baseline, leaving 270,548 participants. We further excluded participants with missing data on selected covariates (including age, sex, race, education, socioeconomic, smoking status, frequency of alcohol consumption, body mass index (BMI), physical activity, and family history of CVD), ultimately leaving 262,343 participants for analyses of the associations between KDM-CBA and various CVDs. Supplementary Figure 1 shows more details.

Construction of KDM-CBA and KDM-CBA Acceleration

In order to accurately and comprehensively portray the biological aging of cardiovascular function, based on previous study,² we screened for classical cardiovascular risk factors, including obesity, dyslipidemia, hypertension, hyperglycemia, and combined them with the biomarkers of biological vascular aging proposed by Hamczyk et al.⁹ (i.e., C-reactive protein (CRP) and insulin-like growth factor 1 [IGF-1]) for a total of 22 candidate biomarkers.

The KDM algorithm is a common method for measuring BA based on composite biomarkers and has been well validated in European and Asian populations and performs well in predicting age-related health outcomes.^6,17 Specifically, the KDM algorithm is used to extract information from biomarkers related to chronological age and to re-predict the estimate of BA.

In this study, we constructed CBA using the KDM. Firstly, for each available biomarker or anthropometric measurement, a Box-Cox transformation was performed to achieve a normal distribution if it was non-normally distributed. Next, 2 biomarkers with a missing rate >30% at baseline were excluded. We then retained only those biomarkers that were significantly correlated with chronological age (correlation coefficient |r| > 0.1 and P<0.05, leaving 9 biomarkers), and based on existing knowledge further analyzed correlations between different biomarkers to exclude redundant biomarkers that might reflect the same aspects of aging (i.e., waist circumference and waist-to-hip ratio (WHR) highly correlated with a correlation coefficient of 0.83, leaving WHR; blood glucose and glycated hemoglobin leaving glycated hemoglobin). Finally, 7 biomarkers remained, comprising body fat percentage (BFP), CRP, glycated hemoglobin, IGF-1, triglycerides, systolic blood pressure, and WHR. The detailed process for the construction of the KDM-CBA is described in the Supplementary Methods and presented in Figure 2.

Figure 2.

Process of constructing KDM-CBA and characteristics of KDM-CBA and KDM-CBA acceleration (N=262,343). (A) Mean (SD) and 95% quartiles for KDM-CBA, KDM-CBA acceleration, and chronological age. (B) Correlation matrix of KDM-CBA and selected biomarkers. (C) Selection of 22 biomarkers and their anthropometric values for the construction of KDM-CBA: final biomarkers shown on light orange background and excluded biomarkers shown on light blue. BFP, body fat percentage; CRP, C-reactive protein; KDM-CBA: Klemera-Doubal method cardiovascular biological age; SBP, systolic blood pressure, WHR, waist-to-hip ratio; TG, triglyceride, HbA1c, glycated hemoglobin; IGF-1, insulin-like growth factor 1; SD, standard deviation.

Furthermore, we measured KDM-CBA acceleration as the difference between KDM-CBA and chronological age to eliminate the interference of age. Positive values of KDM-CBA acceleration may indicate that an individual was older in terms of physiological functioning than expected in the reference sample, and vice versa.

Ascertainment of Outcomes

Based on previous research,^18,19 we included 17 major CVDs, comprising all-cause CVD, coronary atherosclerosis, heart failure, essential hypertension, peripheral artery disease, arrhythmia (atrial fibrillation, cardiac arrest), ischemic heart disease (angina, myocardial infarction and chronic ischemic heart disease) and stroke (hemorrhage stroke, subarachnoid hemorrhage, intracerebral hemorrhage and ischemic stroke). We defined CVD events based on the first occurrence of a set of diagnostic codes for CVD outcomes defined by 3-digit codes from the International Classification of Diseases, 10th edition, as well as the detailed ICD codes for the various CVDs (Supplementary Table 1). Sources of first occurrence data were hospital inpatient records, death registries, primary care records and self-reported health status from UK Biobank assessment clinics. Health data are available for England, Scotland and Wales up to 31 October 2022, 31 August 2022 and 31 May 2022, respectively. Follow-up visits were reviewed at the last date of death, failure of follow-up or availability of information, whichever came first.

Statistical Analysis

We describe the characteristics of the study population, and the data are presented as means and standard deviations for continuous variables and proportions for categorical variables.

In Part I, we constructed the KDM-CBA according to the “Construction of KDM-CBA and KDM-CBA acceleration”. Spearman correlation coefficients were calculated to represent pairwise associations between various composite biomarkers. To validate the effectiveness of CBA, we further explored the associations of CBA and CBA acceleration with all-cause death using the Cox proportional risk model.

In Part II, Cox proportional hazard regression models were used to assess the associations of KDM-CBA and KDM-CBA acceleration with various types of CVD, and the results are reported as hazard ratio (HR) and 95% confidence intervals (95% CI). Model 1 unadjusted; Model 2 adjusted for age and sex; Model 3 further adjusted for race, education, BMI, socioeconomic status (i.e., Townsend Deprivation Index), smoking status, frequency of alcohol consumption, total physical activity, and family history of CVD. Detailed definitions of covariates can be found in Supplementary Table 2. No significant violation of the proportional hazard assumptions was observed when using the Schoenfeld residuals method. In addition, we utilized Cox proportional risk models to compare the associations of chronological age, KDM-CBA and KDM-CBA acceleration with multiple cardiovascular outcomes.

We used restricted cubic spline models with 3 knots (at the 10th, 50th, and 90th percentiles) to assess potential nonlinear associations of KDM-CBA and KDM-CBA acceleration with different types of CVD.

Sensitivity analyses were used to validate the robustness of our findings. Firstly, to account for the competing risk of death, we used the Fine and Gray/subdistribution hazard model, an alternative approach to taking the competing risk of death, to examine the associations of KDM-CBA and KDM-CBA acceleration with various types of CVD. Secondly, we unadjusted BMI to analyze the association of KDM-CBA and KDM-CBA acceleration with different CVD types. Thirdly, considering other potential confounders, we additionally adjusted for blood pressure, blood glucose, and medication history when conducting the sensitivity analyses. Finally, we further excluded cardiovascular events that occurred during the first 2 years of follow-up to rule out reverse causation.

KDM-CBA and KDM-CBA acceleration were demonstrated per standard deviation (SD) increase for comparison. R software (version 4.3.1) was used for statistical analysis and data processing.

Results

Population Characteristics

The characteristics of the participants stratified by incident CVD are summarized in the Table. Among the 262,343 participants, the mean (SD) age was 54.93 (8.13) years, 111,897 (42.7%) were male, and 249,577 (95.1%) were white. 81,628 CVDs developed during an average of 11.43±4.00 years of follow-up. Participants with CVD were more likely to be older, male, smoker, lower socioeconomic status, higher BMI, and have a family history of CVD.

Table.

Characteristics of Study Participants by Incident Cardiovascular Disease

Characteristics	Total participants (N=262,343)	Participants with CVD (N=81,628)	Participants without CVD (N=180,715)
Age at baseline, years	54.93 (8.13)	57.85 (7.73)	53.61 (7.96)
Male (%)	111,897 (42.7)	39,975 (49.0)	71,922 (39.8)
Follow-up, years	11.43 (4.00)	6.83 (3.93)	13.51 (1.56)
White ethnicity (%)	249,577 (95.1)	77,957 (95.5)	171,620 (95.0)
Education (%)
High	99,378 (37.9)	30,869 (37.8)	68,509 (37.9)
Intermediate	100,012 (38.1)	31,246 (38.3)	68,766 (38.1)
Low	17,411 (6.6)	5,388 (6.6)	12,023 (6.7)
Other	45,542 (17.4)	14,125 (17.3)	31,417 (17.4)
SES*	−1.45 (3.00)	−1.33 (3.08)	−1.5 (2.97)
Smoking status (%)
Never	150,294 (57.3)	42,195 (51.7)	108,099 (59.8)
Previous	83,808 (31.9)	29,098 (35.6)	54,710 (30.3)
Current	28,241 (10.8)	10,335 (12.7)	17,906 (9.9)
Frequency of alcohol consumption (%)
Never	18,381 (7.0)	6,558 (8.0)	11,823 (6.5)
<3 times/week	127,726 (48.7)	39,150 (48.0)	88,576 (49.0)
≥3 times/week	116,236 (44.3)	35,920 (44.0)	80,316 (44.4)
Total physical activity (%)
Low	65,371 (24.9)	19,957 (24.4)	45,414 (25.1)
Medium	71,914 (27.4)	20,747 (25.4)	51,167 (28.3)
High	70,519 (26.9)	21,832 (26.7)	48,687 (26.9)
Unknown	54,539 (20.8)	19,092 (23.4)	35,447 (19.6)
BMI, kg/m2	26.60 (4.33)	27.47 (4.61)	26.21 (4.14)
Family history (%)	137,966 (52.6)	46,678 (57.2)	91,288 (50.5)
KDM-CBA	53.94 (11.44)	58.83 (11.01)	51.73 (10.93)
KDM-CBA acceleration	−1.48 (8.31)	0.48 (8.81)	−2.37 (7.92)

*Measured by Townsend deprivation index. BMI, body mass index; CVD, cardiovascular disease; KDM-CBA: Klemera-Doubal method cardiovascular biological age; SES, socioeconomic status.

KDM-CBA and KDM-CBA Acceleration

To assess the agreement between KDM-CBA and chronological age, we displayed scatter plots of KDM-CBA against chronological age, and the results indicated a linear increase in KDM-CBA with chronological age, along with constant variance (mean (SD)=53.94 (11.44) for KDM-CBA) (Figure 2A, Supplementary Figure 2). Negative KDM-CBA acceleration (mean (SD)=−1.48 (8.31)) values can indicate that an individual was younger in terms of physiological functioning than expected for the age of the reference sample. Figure 2A reported that 95% quantiles of KDM-CBA and KDM-CBA acceleration were 31.77–75.26 and −17.73–15.46, respectively. To explore the contribution of the 7 biomarkers to the construction of the KDM-CBA, we created a correlation coefficient matrix plot to illustrate the relationship between KDM-CBA and biomarkers (Figure 2B). The correlation coefficient between CBA and chronological age was 0.69. The correlation coefficients between KDM-CBA and the 7 selected biomarkers ranged from 0.36 to 0.59. The rank order of these correlations was systolic blood pressure, glycated hemoglobin, CRP, IGF-1, BFP, triglyceride and WHR. In the multivariable-adjusted models, an increase of 1 SD in KDM-CBA increased the risk of all-cause death by 29.4% (HR=1.294, 95% CI: 1.261–1.328) (Supplementary Table 3).

Associations of KDM-CBA and KDM-CBA Acceleration With Various Types of CVDs

Both KDM-CBA and KDM-CBA acceleration (per 1SD increase) were associated with increased risk of various CVD in models 1 and 2 (Supplementary Tables 4,5). After adjusting for other confounders (Model 3), these associations remained statistically significant except for atrial fibrillation (Figure 3). In comparison, the associations of KDM-CBA with several CVD outcomes showed stronger associations, including hypertension (HR=2.115, 95% CI: 2.083–2.148), and coronary atherosclerosis (HR=1.711, 95% CI: 1.545–1.896). However, the association between KDM-CBA and atrial fibrillation was not statistically significant (HR=1.017, 95% CI: 0.987–1.047, P=0.274). We observed similar results for KDM-CBA acceleration and KDM-CBA. Compared with chronological age, KDM-CBA and KDM-CBA acceleration showed stronger associations with various CVD (Supplementary Table 6).

Figure 3.

Associations of KDM-CBA and KDM-CBA acceleration with 17 cardiovascular diseases (CVDs) in Part II (N=262,343). The incidences of different types of CVD associated with per 1 SD increase in KDM-CBA and KDM-CBA acceleration are shown. HRs (95% CIs) were calculated in Cox proportional hazards models after adjusting for age, sex, race, body mass index, education, socioeconomic status, smoking status, frequency of alcohol consumption, total physical activity, and family history of CVD. CI, confidence interval; HR, hazard ratio; KDM-CBA: Klemera-Doubal method cardiovascular biological age.

In addition, the dose-response curves of KDM-CBA and KDM-CBA acceleration with different types of CVD are shown in Figure 4 and Supplementary Figure 3 (using restricted cubic spline models). KDM-CBA and KDM-CBA acceleration showed J-type nonlinear associations with nearly all CVD types (cutoff values of ≈55 and −1.7 years for KDM-CBA and KDM-CBA acceleration, respectively).

Figure 4.

Dose-response curves of the KDM-CBA with various cardiovascular diseases (N=262,343). Solid lines: point estimates; Shadows: 95% confidence interval. HR, hazard ratio; KDM-CBA: Klemera-Doubal method cardiovascular biological age.

Sensitivity analyses were robust, and the results were similar in direction and magnitude to those of the main analysis. KDM-CBA and KDM-CBA acceleration were associated with increased risk of developing different CVD types (except for atrial fibrillation) in the Fine and Gray competing models (Supplementary Table 7). The results with unadjusted BMI and further adjusted for blood pressure, blood glucose and medication history were approximately the same as the results of the main analysis (Supplementary Tables 8,9). And the associations between KDM-CBA or KDM-CBA acceleration and different types of CVD remained significant when participants with a follow-up of less than 2 years were excluded (Supplementary Table 10).

Discussion

In this prospective study, which consisted of approximately 260,000 adults from the UK Biobank, the KDM-CBA was constructed via KDM. This study tested and quantified the hypothesis that CBA may be a common starting point of various CVDs not to propose a new CVD predictor and that there may be heterogeneity among these diseases. We observed that both KDM-CBA and KDM-CBA acceleration were associated with the development of different types of CVD, and the strength of the associations varied among different types of CVD. Furthermore, KDM-CBA and KDM-CBA acceleration showed J-type nonlinear associations with nearly all CVD types (cutoff values of ≈55 and −1.7 years for KDM-CBA and KDM-CBA acceleration, respectively). In conclusion, targeting KDM-CBA may serve as a population-level strategy to mitigate the risk of developing diverse CVDs.

Biomarkers such as BFP, CRP, and triglycerides are closely associated with CBA. Firstly, CRP combines with its receptors CD32/CD64, activating the NF-κB signaling pathway to induce the inflammatory process, and promotes aging through the Smad3-dependent p21/p27 mechanism.²⁰ High levels of CRP are associated with increased risk of CVD. Secondly, oxidative stress is a major feature of vascular aging.²¹ Chronic elevation of blood glucose levels stimulates the production of reactive oxygen species and induces oxidative stress,²² thereby promoting vascular aging. Glycated hemoglobin reflects the average blood glucose level over the past few months and is the standard measure of blood glucose control. Regarding BFP and WHR, fat distribution changes with age, with an increase in abdominal fat, which is associated with an increased risk of CVD. High BFP and WHR may be associated with an increased inflammatory response and oxidative stress, thereby accelerating the aging process. In addition, IGF-1 plays an important role in cell growth, proliferation, and repair,²³ and IGF1 signaling is chronically activated during the aging process, resulting in cardiovascular system dysregulation.²³ Cellular senescence is 1 of the 8 characteristics of cardiovascular aging,²³ and lipid overload may promote pathological cardiovascular changes by inducing vascular cell senescence and causing vascular cell dysfunction.²⁴ Finally, hypertension is one of the main risk factors for CVD, and inflammation, oxidative stress, and vascular dysfunction are the underlying mechanisms of both hypertension and BA.²⁵

KDM-CBA and KDM-CBA acceleration emerged as potential indicators of increased incidence of different types of CVD, although different evaluations of different CVD types have been shown. Our findings indicated that KDM-CBA had notably robust associations with hypertension and atherosclerosis, ranking higher than its associations with ischemic heart disease, stroke, and arrhythmia. This discrepancy may be explained by the following underlying mechanisms. Firstly, advancing age contributes to aortic stiffening characterized by increased collagen levels and diminished elastin content, culminating in elevated systolic blood pressure and consequent hypertension.²⁶ Secondly, vascular aging also leads to endothelial dysfunction, reduced antithrombotic properties, increased oxidative stress and inflammatory cytokines, thereby promoting atherosclerosis development.^27,28 Thirdly, the formation of lipid-rich plaques in medium and large arteries, particularly when these plaques develop an unstable phenotype prone to rupture, heightens the risk of stroke occurrence.²⁹ Fourth, progressive cardiomyocyte hypertrophy, inflammation, and cardiac fibrosis are hallmarks of cardiac aging, predisposing individuals to prevalent CVDs such as ischemic heart disease and heart failure.^30,31 Conversely, clinical composite biomarkers may not be able to accurately evaluate the risk of arrhythmogenesis because of cellular-level mechanisms involving ion channels and electrogenic transporters.³² In short, the stronger strength of the association of CBA with hypertension or atherosclerosis may be attributed to the presence of markers directly related to blood pressure and lipids (systolic blood pressure and triglycerides) in the biomarkers used to construct KDM-CBA.

The findings showed no significant correlation between CBA and atrial fibrillation (AF), which we attempted to explore in terms of the mechanisms of AF occurrence and the construction of CBA. The mechanisms of AF are complex and involve the interaction of multiple factors.³³ Firstly, AF is often associated with electrophysiological alterations, including abnormal impulse conduction (i.e., tissue fibrosis and other extracellular matrix changes in the myocardium, connexin dysregulation, and Na⁺-channel dysfunction), electrical refractoriness, and impulse generation.³⁴ Additionally, conduction abnormalities are associated with atrial structural remodeling, primarily involving tissue fibrosis.³⁵ As far as the construction of CBA is concerned, this study based the construction on the composite clinical biomarkers of blood measurements and physical measurements, extracting information related to cardiovascular aging as much as possible from the perspective of physical blood biochemical indicators without including electrophysiological information. Although electrophysiological abnormalities, as well as changes in atrial structure, may be age-related, CBA may not fully capture these specific electrophysiological changes and may not directly reflect subtle changes in atrial structure.

Aging is a dynamic and nonlinear process.³⁶ Our findings revealed nonlinear associations between KDM-CBA and the occurrence of different CVDs (hypertension, heart failure, arrhythmia, ischemic heart disease, etc.). Notably, a CBA of 55 years and CBA acceleration of −1.7 may serve as potential cutoff values. A study has indicated that the age of 55 years could be a crucial turning point in life, with approximately 20% of women experiencing a substantial reduction in their health at that age, during which time ovarian function deteriorates, and endocrine levels decline significantly.³⁷ Our findings offer valuable insights into early CVD prevention, treatment, or high-risk group identification based on CBA. Similarly, previous studies have characterized the nonlinear associations of aging and significant nonlinear changes in the plasma proteome.³⁶ Consistent with our observations, some studies have demonstrated an enrichment of CVD-associated proteins in the proteomic profiles of middle-aged and older adults compared with younger counterparts, aligning with the increased CVD incidence in older age groups.^36,38

We propose that KDM-CBA serves as both a composite indicator and an early intervention measure suitable for early routine monitoring.^39,40 The proposal regarding KDM-CBA aligns closely with the paradigm of comprehensive management. In the past, the single management approach may have only required the control of blood pressure or the control of blood glucose, but now the aggregation of all the biomarkers related to CVD corresponds to a clearer comprehensive intervention target. Second, for outcomes, if blood pressure is controlled alone, it may only affect the distal outcome of abnormal blood pressure, and KDM-CBA is a composite indicator that can be associated with almost all CVD types. Controlling CBA from multiple biomarker dimensions achieves a more integrated management result, and there is no additional cost for data derived from physical examination records.¹⁴

Study Strengths and Limitations

The strength of this study lies in the comprehensive exploration of the associations between CBA based on composite biomarkers reflecting biological aging of cardiovascular function and various types of CVD, offering unique insights to inform clinical interventions targeting cardiovascular aging using composite biomarkers. However, some limitations should be acknowledged when interpreting the results. First, the different CVD types may be closely and universally correlated, but due to the lack of a well-established and widely applied approach, we did not consider associations between different types in the present study. Second, UK Biobank represents a volunteer cohort, which may include healthier individuals compared with the general population, and the majority of participants were white, thus limiting the generalizability of the results, which should be extrapolated with caution. Third, due to data accessibility, we could not cover all clinical composite biomarkers associated with aging, and therefore constructed CBA to capture only specific aspects of the cardiovascular biological aging process. In addition, the currently constructed CBA is mainly based on blood biochemical biomarkers, capturing only specific aspects of the cardiovascular biological aging process. It is difficult to effectively reflect abnormalities associated with electrophysiology. Fourth, given that constructing a CBA relies on clinical biomarkers and UK Biobank data accessibility, to ensure as representative a sample as possible and statistical validity, clinical biomarkers were measured only at baseline for the entire population. Fifth, although we have made efforts to control for all potential confounders, it is important to consider that there may be other unmeasured confounders that will have to be further explored in future studies.

Conclusions

In summary, our study tested and quantified the hypothesis that CBA may be a common starting point of various CVDs, and that there may be heterogeneity among these diseases. These findings highlight the significance of CBA as an early biomarker for CVD, offering valuable insights for CVD interventions and having important implications for early clinical diagnosis and public health policy guidance.

Ethics Approval

The UK Biobank study received approval from the North West Multi-Center Research Ethics Committee. All participants signed the informed consent before participating in this survey. This research was conducted using UK Biobank Resource under Application Number 117185. We thank all team members and participants involved in the UK Biobank.

Authors’ Contributions

J.C., R.Y.: Research design, Data analysis, Writing- review & editing. N.Z.: Data analysis, Review & editing. H.Z., Y.Z., Y.X., H.X.: Review & editing. X.Z.: Review & editing, Supervision. X.X.: Validation, Supervision, Writing- review & editing, Project administration, Funding acquisition.

Funding

This work was supported and funded by the National Natural Science Foundation of China (Grant No. 82273740) and Sichuan Science and Technology Program (Natural Science Foundation of Sichuan Province, Grant No. 2024NSFSC0552).

Data Availability

UK Biobank data are available online (https://www.ukbiobank.ac.uk).

Consent for Publication

Not applicable.

Conflict of Interest

We declare no competing interests.

Disclosure Summary

The authors have nothing to disclose.

Supplementary Files

Please find supplementary file(s);

https://doi.org/10.1253/circj.CJ-24-0824

References

• 1.   Virani SS, Alonso A, Aparicio HJ, Benjamin EJ, Bittencourt MS, Callaway CW, et al. Heart disease and stroke statistics: 2021 update: A report from the American Heart Association. Circulation 2021; 143: e254–e743, doi:10.1161/cir.0000000000000950.
• 2.   Nie C, Li Y, Li R, Yan Y, Zhang D, Li T, et al. Distinct biological ages of organs and systems identified from a multi-omics study. Cell Rep 2022; 38: 110459, doi:10.1016/j.celrep.2022.110459.
• 3.   Zhang N, Liu X, Wang L, Zhang Y, Xiang Y, Cai J, et al. Lifestyle factors and their relative contributions to longitudinal progression of cardio-renal-metabolic multimorbidity: A prospective cohort study. Cardiovasc Diabetol 2024; 23: 265, doi:10.1186/s12933-024-02347-3.
• 4.   Fujiyoshi A, Kohsaka S, Hata J, Hara M, Kai H, Masuda D, et al. JCS 2023 Guideline on the primary prevention of coronary artery disease. Circ J 2024; 88: 763–842, doi:10.1253/circj.CJ-23-0285.
• 5.   Belsky DW, Caspi A, Houts R, Cohen HJ, Corcoran DL, Danese A, et al. Quantification of biological aging in young adults. Proc Natl Acad Sci USA 2015; 112: E4104–E4110, doi:10.1073/pnas.1506264112.
• 6.   Levine ME. Modeling the rate of senescence: Can estimated biological age predict mortality more accurately than chronological age? J Gerontol A Biol Sci Med Sci 2013; 68: 667–674, doi:10.1093/gerona/gls233.
• 7.   Jylhävä J, Pedersen NL, Hägg S. Biological age predictors. EBioMed 2017; 21: 29–36, doi:10.1016/j.ebiom.2017.03.046.
• 8.   Bersani FS, Canevelli M, Cesari M, Maggioni E, Pasquini M, Wolkowitz OM, et al. Frailty Index as a clinical measure of biological age in psychiatry. J Affect Dis 2020; 268: 183–187, doi:10.1016/j.jad.2020.03.015.
• 9.   Hamczyk MR, Nevado RM, Barettino A, Fuster V, Andrés V. Biological versus chronological aging: JACC Focus Seminar. J Am Coll Cardiol 2020; 75: 919–930, doi:10.1016/j.jacc.2019.11.062.
• 10.   Tian YE, Cropley V, Maier AB, Lautenschlager NT, Breakspear M, Zalesky A. Heterogeneous aging across multiple organ systems and prediction of chronic disease and mortality. Nat Med 2023; 29: 1221–1231, doi:10.1038/s41591-023-02296-6.
• 11.   Childs BG, Durik M, Baker DJ, van Deursen JM. Cellular senescence in aging and age-related disease: From mechanisms to therapy. Nat Med 2015; 21: 1424–1435, doi:10.1038/nm.4000.
• 12.   Chen L, Zhang Y, Yu C, Guo Y, Sun D, Pang Y, et al. Modeling biological age using blood biomarkers and physical measurements in Chinese adults. EBioMed 2023; 89: 104458, doi:10.1016/j.ebiom.2023.104458.
• 13.   Bruno RM, Nilsson PM, Engström G, Wadström BN, Empana JP, Boutouyrie P, et al. Early and supernormal vascular aging: Clinical characteristics and association with incident cardiovascular events. Hypertension 2020; 76: 1616–1624, doi:10.1161/hypertensionaha.120.14971.
• 14.   Kang YG, Suh E, Lee JW, Kim DW, Cho KH, Bae CY. Biological age as a health index for mortality and major age-related disease incidence in Koreans: National Health Insurance Service: Health screening 11-year follow-up study. Clin Interv Aging 2018; 13: 429–436, doi:10.2147/cia.S157014.
• 15.   Chen MS, Lee RT, Garbern JC. Senescence mechanisms and targets in the heart. Cardiovasc Res 2022; 118: 1173–1187, doi:10.1093/cvr/cvab161.
• 16.   North BJ, Sinclair DA. The intersection between aging and cardiovascular disease. Circ Res 2012; 110: 1097–1108, doi:10.1161/circresaha.111.246876.
• 17.   Chan MS, Arnold M, Offer A, Hammami I, Mafham M, Armitage J, et al. A biomarker-based biological age in UK Biobank: Composition and prediction of mortality and hospital admissions. J Gerontol A Biol Sci Med Sci 2021; 76: 1295–1302, doi:10.1093/gerona/glab069.
• 18.   Ai S, Zhang J, Zhao G, Wang N, Li G, So HC, et al. Causal associations of short and long sleep durations with 12 cardiovascular diseases: Linear and nonlinear Mendelian randomization analyses in UK Biobank. Eur Heart J 2021; 42: 3349–3357, doi:10.1093/eurheartj/ehab170.
• 19.   Schuermans A, Vlasschaert C, Nauffal V, Cho SMJ, Uddin MM, Nakao T, et al. Clonal haematopoiesis of indeterminate potential predicts incident cardiac arrhythmias. Eur Heart J 2024; 45: 791–805, doi:10.1093/eurheartj/ehad670.
• 20.   Tang Y, Fung E, Xu A, Lan HY. C-reactive protein and ageing. Clin Exp Pharmacol Physiol 2017; 44(Suppl 1): 9–14, doi:10.1111/1440-1681.12758.
• 21.   Ding YN, Tang X, Chen HZ, Liu DP. Epigenetic regulation of vascular aging and age-related vascular diseases. Adv Exp Med Biol 2018; 1086: 55–75, doi:10.1007/978-981-13-1117-8_4.
• 22.   Bhansali S, Bhansali A, Walia R, Saikia UN, Dhawan V. Alterations in mitochondrial oxidative stress and mitophagy in subjects with prediabetes and type 2 diabetes mellitus. Front Endocrinol 2017; 8: 347, doi:10.3389/fendo.2017.00347.
• 23.   Abdellatif M, Rainer PP, Sedej S, Kroemer G. Hallmarks of cardiovascular ageing. Nat Rev Cardiol 2023; 20: 754–777, doi:10.1038/s41569-023-00881-3.
• 24.   Zeng Q, Gong Y, Zhu N, Shi Y, Zhang C, Qin L. Lipids and lipid metabolism in cellular senescence: Emerging targets for age-related diseases. Ageing Res Rev 2024; 97: 102294, doi:10.1016/j.arr.2024.102294.
• 25.   Buford TW. Hypertension and aging. Ageing Res Rev 2016; 26: 96–111, doi:10.1016/j.arr.2016.01.007.
• 26.   Zieman SJ, Melenovsky V, Kass DA. Mechanisms, pathophysiology, and therapy of arterial stiffness. Arterioscler Thromb Vasc Biol 2005; 25: 932–943, doi:10.1161/01.Atv.0000160548.78317.29.
• 27.   Gimbrone MA Jr, García-Cardeña G. Endothelial cell dysfunction and the pathobiology of atherosclerosis. Circ Res 2016; 118: 620–636, doi:10.1161/circresaha.115.306301.
• 28.   Donato AJ, Eskurza I, Silver AE, Levy AS, Pierce GL, Gates PE, et al. Direct evidence of endothelial oxidative stress with aging in humans: Relation to impaired endothelium-dependent dilation and upregulation of nuclear factor-kappaB. Circ Res 2007; 100: 1659–1666, doi:10.1161/01.RES.0000269183.13937.e8.
• 29.   Puylaert P, Zurek M, Rayner KJ, De Meyer GRY, Martinet W. Regulated necrosis in atherosclerosis. Arterioscler Thromb Vasc Biol 2022; 42: 1283–1306, doi:10.1161/atvbaha.122.318177.
• 30.   Meschiari CA, Ero OK, Pan H, Finkel T, Lindsey ML. The impact of aging on cardiac extracellular matrix. GeroScience 2017; 39: 7–18, doi:10.1007/s11357-017-9959-9.
• 31.   Lakatta EG, Levy D. Arterial and cardiac aging: Major shareholders in cardiovascular disease enterprises. Part II: The aging heart in health: Links to heart disease. Circulation 2003; 107: 346–354, doi:10.1161/01.cir.0000048893.62841.f7.
• 32.   Delisle BP, Anson BD, Rajamani S, January CT. Biology of cardiac arrhythmias: Ion channel protein trafficking. Circ Res 2004; 94: 1418–1428, doi:10.1161/01.RES.0000128561.28701.ea.
• 33.   Takase B, Ikeda T, Shimizu W, Abe H, Aiba T, Chinushi M, et al. JCS/JHRS 2022 Guideline on diagnosis and risk assessment of arrhythmia. Circ J 2024; 88: 1509–1595, doi:10.1253/circj.CJ-22-0827.
• 34.   Nattel S, Heijman J, Zhou L, Dobrev D. Molecular basis of atrial fibrillation pathophysiology and therapy: A translational perspective. Circ Res 2020; 127: 51–72, doi:10.1161/circresaha.120.316363.
• 35.   Andrade J, Khairy P, Dobrev D, Nattel S. The clinical profile and pathophysiology of atrial fibrillation: Relationships among clinical features, epidemiology, and mechanisms. Circ Res 2014; 114: 1453–1468, doi:10.1161/circresaha.114.303211.
• 36.   Lehallier B, Gate D, Schaum N, Nanasi T, Lee SE, Yousef H, et al. Undulating changes in human plasma proteome profiles across the lifespan. Nat Med 2019; 25: 1843–1850, doi:10.1038/s41591-019-0673-2.
• 37.   Solomon DH, Colvin A, Lange-Maia BS, Derby C, Dugan S, Jackson EA, et al. Factors associated with 10-year declines in physical health and function among women during midlife. JAMA Netw Open 2022; 5: e2142773, doi:10.1001/jamanetworkopen.2021.42773.
• 38.   Go AS, Mozaffarian D, Roger VL, Benjamin EJ, Berry JD, Borden WB, et al. Heart disease and stroke statistics: 2013 update: A report from the American Heart Association. Circulation 2013; 127: e6–e245, doi:10.1161/CIR.0b013e31828124ad.
• 39.   Jiang M, Tian S, Liu S, Wang Y, Guo X, Huang T, et al. Accelerated biological aging elevates the risk of cardiometabolic multimorbidity and mortality. Nat Cardiovasc Res 2024; 3: 332–342, doi:10.1038/s44161-024-00438-8.
• 40.   Gao X, Geng T, Jiang M, Huang N, Zheng Y, Belsky DW, et al. Accelerated biological aging and risk of depression and anxiety: Evidence from 424,299 UK Biobank participants. Nat Commun 2023; 14: 2277, doi:10.1038/s41467-023-38013-7.