Skip to main content
Download PDF

Development and validation of a five-year cardiovascular risk assessment tool for Asian adults aged 75 years and older

BMC Geriatrics volume 25, Article number: 15 (2025) Cite this article

Abstract

Background

To identify cardiovascular (CV) risk factors in Asian elderly aged 75 years and older and subsequently develop and validate a sex-specific five-year CV risk assessment tool for this population.

Methods

This study included 12,174 patients aged ≥ 75 years without a prior history of cardiovascular disease at a single hospital in Taiwan. Electronic health records were linked to the National Health Insurance Research Database and the National Death Registry to ensure comprehensive health information. Eligible patients were randomly divided into derivation (80%) and validation (20%) cohorts. A sex-specific CV risk assessment tool was developed to predict major adverse cardiovascular events (MACE) using Cox regression modeling.

Results

During a median follow-up period of 8.6 years for men and 8.5 years for women in the derivation cohort, MACE occurred in 3.62% of men and 3.02% of women. Predictors for men comprised advanced age, smoking, non-HDL-C levels > 160 mg/dL, metastatic cancer, and aspirin usage. Predictors for women included advanced age, smoking, atrial fibrillation, cancer, dementia, osteoarthritis, systemic lupus erythematosus, use of antihypertensives, and use of oral anticoagulants. In the validation cohort, the sex-specific risk assessment tool demonstrated fair discriminative power (AUC: men, 0.64; women, 0.68). Model calibration demonstrated good performance for women but was less optimal for men.

Conclusions

This sex-specific CV risk assessment tool shows fair discriminative capability in estimating risk of cardiovascular disease among elderly Asians, potentially enabling targeted interventions in this vulnerable population.

Peer Review reports

Background

Cardiovascular disease (CVD) continues to exert a significant toll on the well-being of older adults and represents a leading cause of mortality and disability [1]. As the elderly population expands significantly, it becomes increasingly imperative to proactively manage cardiovascular (CV) risk factors. Employing risk assessment tools in this context holds the potential for significant benefits in primary CVD prevention. These tools play a pivotal role in the early identification of high-risk individuals, thereby facilitating the implementation of targeted preventive interventions.

Several CVD risk assessment tools, such as the Pooled Cohort Equations (PCE), Framingham CVD Risk Score, Systematic Coronary Risk Evaluation 2 (SCORE2), and QRESEARCH Cardiovascular Risk Algorithm 3 (QRISK3), have been endorsed in primary prevention guidelines for estimating absolute CV risk [2,3,4]. However, these tools often manifest suboptimal performance when applied to older adults. A notable example is the widely utilized PCE, which has a tendency to overestimate CV risk in elderly individuals [5]. The intricacies surrounding CV risk prediction in older adults arise from their diverse health profiles, including frailty, cognitive decline, multiple coexisting health conditions, and altered physiological responses to traditional CV risk factors [6, 7]. Moreover, traditional CV risk factors, such as blood pressure, may exhibit different associations with CVD in older individuals [8]. The use of preventive medications for CVD in older adults also presents a challenge, as their benefits and potential side effects demand careful consideration [9]. Consequently, the risk factors and scoring systems underpinning existing assessment tools primarily derived from middle-aged populations may inadequately capture the complexity of CV risk assessment in the elderly.

Majority of the assessment tool developments have been limited to Western populations [10,11,12,13]. Among them, the SCORE2-Older Persons (SCORE2-OP) model was specifically derived for individuals aged 70 years and older; however, its development relied heavily on data from the Cohort of Norway (CONOR) study, which predominantly included European participants [13]. Racial and ethnic disparities in CVD risk factors may limit the applicability of such tools in non-Western populations. For example, studies have reported overestimation of CV risk in Asian populations when using the PCE and SCORE2-OP models [14, 15]. While prediction models derived from Asian data exist, they primarily target middle-aged patients, with limited representation of elderly populations [16,17,18]. As life expectancy continues to increase, there is a compelling need to develop a CV risk assessment tool tailored to Asians, particularly those aged 75 years and older.

Additionally, considering the potential differences in the prevalence and impact of CV risk factors between males and females [19,20,21,22], it is pertinent to develop distinct prediction models for each sex. Recognizing these gaps, the goal of this study was to identify CV risk factors in Asian elderly and subsequently develop and validate a sex-specific five-year CV risk assessment tool for those aged 75 years and older, thereby enhancing clinical care for this population.

Methods

Study design and data source

This study utilized a subset of data from the National Taiwan University Hospital-integrative Medical Database (NTUH-iMD) which included patients aged 20 years and older diagnosed with hypertension, diabetes mellitus, dyslipidemia, or any CVD between January 2006 and July 2017 at NTUH. The NTUH-iMD is a comprehensive electronic health record database containing patient demographics, diagnostic records, laboratory results, procedural data, and pharmacy records from a single medical center. The NTUH-iMD subset, with data available from 2006 to 2017, was linked to the 2003–2017 National Health Insurance Research Database (NHIRD), Cause of Death Registry, and Catastrophic Illness File in Taiwan. The NHIRD is a national administrative claims database that provides patient-level information on demographics, diagnoses, outpatient and inpatient service procedures, prescriptions, and enrollment data for individuals under the coverage of Taiwan's National Health Insurance (NHI) [23]. The NHI is a single-payer compulsory health insurance system that covers over 99% of the Taiwanese population. The Cause of Death Registry documents the date and cause of death of NHI program enrollees. In this study, the Catastrophic Illness File was employed to ascertain the dialysis status of patients.

The study protocol was approved by the Research Ethics Committee of NTUH (No. 201710009RINC), and informed consent was not required for this study.

Study population

The study cohort consisted of individuals aged 75 years and older who visited NTUH between February 2008 and December 2012. Cohort entry date was defined as the patient’s first visit date to NTUH during this period after reaching the age of 75. Exclusions were applied to patients with any diagnosis of CVD, which included angina, myocardial infarction (MI), stroke, transient ischemic attack, heart failure, peripheral artery disease, and valvular heart disease, before their cohort entry date. This exclusion criterion was applied using a minimum five-year lookback period prior to the cohort entry date to ensure accurate identification of incident cases. Detailed International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] diagnosis codes are provided in Supplementary Table S1. Patients with a history of percutaneous coronary intervention or coronary artery bypass graft were excluded as well, with the criteria determined using the International Classification of Diseases, Ninth or Tenth Revision, Procedure Coding System (ICD-9-PCS or ICD-10-PCS) codes listed in Supplementary Table S2. Patients who were disenrolled from the NHI program before their cohort entry date were also excluded.

Outcomes and follow-up

The study aimed to investigate major adverse cardiovascular events (MACE), defined as a composite outcome comprising non-fatal myocardial infarction, non-fatal ischemic stroke, and cardiovascular death. The occurrence of MACE was determined by identifying relevant diagnosis codes in the primary diagnosis of hospitalization records in the NHIRD or relevant death records in the Cause of Death Registry. We used both ICD-9-CM and ICD-10-CM diagnosis codes, taking into account the transition from ICD-9-CM to ICD-10-CM in 2016. Specific diagnosis codes for defining MACE outcomes are detailed in Supplementary Table S3.

The start of the outcome follow-up (the index date) was defined as 180 days after the cohort entry date. This 180-day lag time served two purposes: it allowed for the confirmatory exclusion of any history of CVD and enabled the collection of laboratory data for new patients at NTUH. Patients were followed from the index date to the date of MACE occurrence, non-CV death, disenrollment from the NHI program, or the end of the study period on December 31, 2017.

Measurement of risk factors

Thirty candidate predictor variables were selected for analysis based on existing literature, clinician expertise, and the data available in our databases (Supplementary Table S4). These variables included age, sex, body mass index (BMI), smoking history, and laboratory results (low-density lipoprotein cholesterol [LDL-C], high-density lipoprotein cholesterol [HDL-C], total cholesterol [T-CHO], and triglycerides levels [24]), which were extracted from the NTUH-iMD. Laboratory data reported within one year before the index date was taken into consideration, and for each patient, the measurement closest to the index date was utilized. Non-HDL-C values were calculated using the formula: non-HDL-C = T-CHO – HDL-C. In addition to these factors, our analysis incorporated the presence of baseline comorbidities, such as hypertension, diabetes, atrial fibrillation, chronic obstructive pulmonary disease (COPD), cancer, chronic kidney disease, end-stage renal disease requiring dialysis, dementia, and autoimmune disease. Baseline comorbidities were identified based on either an inpatient discharge diagnosis or at least two outpatient diagnoses within two years preceding the index date. We calculated the multimorbidity frailty index (mFI) to assess the degree of frailty by considering specific conditions for each patient [25]. Furthermore, medication use at baseline was determined by having at least two corresponding prescriptions within one year before the index date. The complete list of diagnosis codes and Anatomical Therapeutic Chemical (ATC) codes for medications is provided in Supplementary Table S4.

Continuous variables underwent logarithm transformation to enhance normality and homoscedasticity, mitigating the impact of extreme values [17]. Specifically, age, BMI, and lipid profiles were evaluated both as categorical variables based on their original continuous form and as continuous variables following natural logarithm transformation. The categorical form of the variables was selected for clinical convenience if the model utilizing log-transformed continuous variables did not substantially improve model performance.

Statistical analysis

Descriptive statistics are presented as mean ± standard deviation (SD) for continuous variables and as number and percentage for categorical variables. A P-value of < 0.05 was considered statistically significant. The independent t-test was used to compare means between men and women, while the Chi-square test was employed to compare distributions for categorical variables.

Model development

Five-year CVD risk was defined as the probability of experiencing the first MACE event within a five-year period for individuals without history of CVD at baseline. Sex-specific Cox proportional hazards models were utilized to develop CVD risk prediction equations. The study cohort for each sex was randomly divided into two subsets, with 80% allocated for model derivation and 20% for model validation. Initial bivariate analysis assessed the unadjusted odds ratio of each candidate risk factor for the first occurrence of MACE, excluding variables with a P-value ≥ 0.25 from subsequent multivariable regression. Variables showing significant interaction with age (P-value < 0.05) were integrated as interaction terms. The final Cox proportional hazards model was developed using a stepwise selection approach, combining both forward (P-value < 0.25) and backward (P-value > 0.15) criteria to minimize the Akaike Information Criterion (AIC). The model with the lowest AIC, indicative of good model fitness and reduced overfitting risk, was selected. Cox proportional hazards models estimated coefficients and 95% confidence intervals for hazard ratios associated with each risk factor. All statistical analyses were performed using SAS software (version 9.4, SAS Institute Inc., Cary, NC, USA).

In the primary analysis, multiple imputations were employed to address missing values in smoking history, BMI, and lipid profiles, which had approximately 20%, 42%, and 59 ~ 68% missing data, respectively. This approach imputed missing values based on the observed relationships within the dataset, mitigating the limitations associated with single-value imputation such as reducing variances in data distribution [26]. Five imputed datasets were generated through Markov Chain Monte Carlo simulation in SAS 9.4, and the results from the five Cox models were subsequently combined using PROC MIANALYZE.

Evaluation of model performance

We assessed the performance of our prediction models through two key measures: discrimination and calibration [27, 28]. Discrimination was evaluated using Harrell's concordance index (C-index), an adapted metric from the area under the receiver operating curve (AUC; also called C statistic) for survival analysis [29]. An AUC or C-index value below 0.6 was considered poor quality, values from 0.6 to 0.7 indicated fair quality, and values above 0.7 represented good quality [30]. Calibration was examined with the D'Agostino-Nam test, an extension of the Hosmer–Lemeshow goodness-of-fit test for survival data, which divided the data into risk groups to determine the alignment of predicted risks with observed risks, with a non-significant P-value indicating good calibration [31]. With the proportion of censored data meeting the prerequisites for the D'Agostino-Nam test (15% for male and 10% for female) [32], we calculated observed risks through Kaplan–Meier estimates at the five-year mark.

Results

Patient characteristics and incidence of CVD

A total of 42,061 patients aged 75 or older who visited NTUH between February 2008 and December 2012 were initially identified from the NTUH-iMD. After excluding patients with disenrollment from NHI or a history of CVD at baseline, 12,174 eligible patients remained in the study cohort (Fig. 1). The derivation cohort used for model development consisted of a random 80% sample of the overall cohort, including 4,358 men and 5,382 women. The remaining 20% of the overall cohort was used for model validation. In both the derivation and validation cohorts, the mean age at baseline was 79 years for both sexes (Table 1). Baseline BMI and lipid profiles before and after multiple imputations are provided in Supplementary Table S5. Men had a higher prevalence of smoking and COPD, women had a higher prevalence of hypertension, and lipid levels in women were statistically higher compared to men. The baseline characteristics of the derivation and validation cohorts were generally similar.

Fig. 1
figure 1

Study flow chart. Abbreviations: CABG, coronary artery bypass graft; CVD, cardiovascular disease; NHI, National Health Insurance; NTUH, National Taiwan University Hospital; PCI, percutaneous coronary intervention

Full size image
Table 1 Baseline characteristics of included cohorts
Full size table

In the derivation cohort, the median follow-up duration was 8.6 years for men and 8.5 years for women. During the follow-up period, MACE occurred in 3.62% (590 events/162,698 person-years [PYs]) of men and 3.02% (608 events/201,463 PYs) of women. The primary contributors to MACE were ischemic stroke (56%), followed by CV death (36%), and MI (9%) (Supplementary Table S6). Within the validation cohort, both sexes had a median follow-up of 7.5 years; MACE occurred in 3.55% (144 events/40,606 PYs) of men and 3.26% (163 events/49,999 PYs) of women during the follow-up period. The predominant causes of MACE were consistent with those identified in the derivation cohort (Supplementary Table S6).

The bivariate analysis considered all candidate predictor variables. Among men, 23 out of 30 candidate predictors showed potential association with MACE (P-value < 0.25), encompassing factors such as age, BMI, lipid profiles, smoking, frailty, most comorbidities, and some baseline medications (as shown in Supplementary Table S7). In the case of women, 21 out of 30 candidate predictors were identified as potential factors associated with MACE, including older age, BMI, HDL-C, smoking, most comorbidities, and most baseline medications (Supplementary Table S8).

Sex-specific CV risk assessment tool

To optimize model performance and clinical applicability, the final model with categorical variables was selected. A comparison of model performance between categorical and continuous models is presented in Supplementary Table S9. Stepwise multivariable regression identified five statistically significant predictor variables associated with MACE in men and nine variables in women. In men, significant predictor variables included advanced age, smoking, non-HDL-C levels exceeding 160 mg/dL, metastatic cancer, and previous aspirin use (Table 2). The significant factors in women included advanced age, smoking, atrial fibrillation, cancer, systemic lupus erythematosus, dementia, osteoarthritis, and prior use of oral anticoagulants and antihypertensive medications. Notably, none of the lipid profiles were found to be significantly linked to MACE, and an interaction was observed between age and antihypertensive medications (Table 3).

Table 2 Predictive model for estimation of five-year cardiovascular risk in men
Full size table
Table 3 Predictive model for estimation of five-year cardiovascular risk in women
Full size table

Model performance

The final model demonstrated fair discrimination in predicting MACE risk in the derivation cohort, with a C-index of 0.63 in men and 0.67 in women (Table 4). Calibration results indicated good model fit for women but weaker calibration for men (P-value: 0.36 in women, < 0.05 in men, Table 4). Figure 2 depicts the comparison between predicted and observed CV risk across deciles, indicating a slight overestimation of CV risk in the model for men. During internal validation, the model showed slight improvement in discrimination (C-index: 0.64 in men and 0.68 in women) while maintaining similar calibration results (Hosmer–Lemeshow tests, P-value: 0.26 in women, < 0.05 in men) (Table 4). CV risk for men remained slightly overestimated across deciles (Fig. 2).

Table 4 Discrimination and calibration performance of prediction models
Full size table
Fig. 2
figure 2

Calibration plots of models: A men-derivation cohort (B) men-validation cohort (C) women-derivation cohort (D) women-validation cohort. Abbreviation: CVD, cardiovascular disease

Full size image

Discussion

In our study, we identified CV risk factors and developed a sex-specific five-year CV risk prediction model for individuals 75 years of age and older in Taiwan. Our models exhibited fair discrimination for both sexes and good calibration for women. To the best of our knowledge, it represents the first sex-specific CV risk assessment tool for older Asian adults in recent decades [33]. The landscape of CVD, its risk factors, and management have substantially evolved since previous prediction models were developed [34, 35]. By utilizing contemporary data for model development, this tool has the potential to improve the accuracy of risk assessment in elderly Asians, enabling more precise identification of high-risk individuals and optimizing patient selection for preventive medications.

To compile comprehensive data, we linked the NTUH-iMD with the NHIRD, creating a robust information source that includes laboratory results and mortality records. Our study also considered important medications, encompassing aspirin, antihypertensive drugs, antihyperglycemic drugs, anticoagulants, and statins. Unlike many existing assessment tools that primarily focus on antihypertensive drugs, ours offers a more comprehensive approach that aligns with contemporary guidelines. Moreover, we addressed the intricacies of CV risk estimation in the elderly by incorporating frailty and dementia as potential risk factors. Beyond traditional risk factors, our research highlights the importance of chronic comorbidities such as dementia and medications such as antihypertensive drugs, aspirin, and oral anticoagulants in assessing CV risk in the elderly.

Our study aligns with previous research that highlighted the diminishing significance of LDL-C with advancing age and the more limited benefit of lowering LDL-C in reducing CV risk in elderly individuals. With the exception of non-HDL-C in men, our results suggest that other lipid parameters, including LDL-C, T-CHO, and TG, did not emerge as robust markers for CV risk prediction. This is consistent with the findings that non-HDL-C outperformed LDL-C in predicting CV events [36]. In the era of statin treatment, non-HDL-C levels, rather than LDL-C, have been identified as the most significant residual predictor of MACE risk among the lipid profile in patients with diabetes mellitus [37]. Moreover, previous studies have shown that the impact of LDL-C decreases as individuals age [38], and older adults tend to exhibit lower LDL-C levels upon admission for acute MI, suggesting that LDL-C may not have a causative link with CVD in this population [39]. When considering the use of statins in primary prevention for individuals aged 75 and above, the existing body of evidence remains limited [40]. To definitively establish the potential benefits of various lipid-lowering therapies in this demographic, larger-scaled studies are warranted.

To account for the data limitations in our database, we utilized hypertension diagnosis and use of antihypertensive medications as proxies for high systolic blood pressure, a conventional CV risk factor. Despite the inherent challenge in accurately assessing individual blood pressure control, our analysis did not identify the use of antihypertensive drugs as a significant predictor for older men, but it remained significant for women. This finding partially aligns with the results of the QRISK2 study, which investigated the relationship between systolic blood pressure and CVD across various age groups, noting a diminishing impact of blood pressure with aging [41]. Similarly, a study conducted in the Netherlands excluded systolic blood pressure when calculating CVD risk in older adults [42]. It is important to note that 70% of our study population had hypertension diagnosis and took antihypertensives. Consequently, the influence of blood pressure on CVD risk in men likely waned, leading to its exclusion during the model stepwise selection process. In contrast, as reported in the literature, hypertension becomes more prevalent in elderly women [43]. Some researchers suggest that menopausal women experience an increased prevalence of high blood pressure because estrogen has a protective effect on inhibiting inflammation and progression of atherosclerosis [44]. However, despite more than 70% of women with hypertension being aware of their diagnosis and receiving treatment, less than one-third achieve adequate blood pressure control [45]. The reasons for inadequate blood pressure control in older women remain unclear, with potential factors including inadequate treatment intensity, inappropriate drug choices, and lack of adherence. These factors may explain why the use of antihypertensive drugs remained a significant risk factor for MACE in women.

In comparison to established risk prediction models for the elderly, our model exhibited a slightly lower C-index than the short-term PCE with the addition of three biomarkers (AUC 0.75) but demonstrated similar performance to the short-term PCE (AUC 0.65) and SCORE2-OP (AUC 0.66) as reported in prior studies [10, 13]. However, direct comparison of the C-index between these models may not be appropriate due to their study populations and design. The short-term PCE is not a sex-specific risk assessment tool, whereas our model was specifically developed to account for sex differences in cardiovascular risk. Additionally, both PCE and SCORE-OP incorporate variables, such as systolic blood pressure, that were not available in our dataset, further limiting the feasibility of a direct evaluation of their comparative predictive performance in our study.

Our study has several limitations worth noting. First, the use of secondary health databases introduces inherent limitations, including the potential for misclassification in claims data and missing data in electronic health records. To mitigate the risk of misclassification, we extended the baseline look-back period to a minimum of five years to exclude patients with a history of CVD. Second, our model was developed using a cohort consisting of over 70% of patients with hypertension, diabetes, or dyslipidemia. External validation in general or more diverse populations is needed to evaluate its performance and applicability in various settings. Third, we did not apply a competing risk-adjusted model, which may have contributed to overestimation of CV risk, particularly in low-risk patients, as non-CVD deaths could preclude the occurrence of CVD events. The overestimation of CV risk may also reflect the influence of unmeasured confounders, such as frailty or lifestyle factors, and the disproportionate effect of high-risk individuals on the model coefficients. Fourth, our database lacked critical information on traditional risk factors such as systolic blood pressure and family history, and emerging biomarkers such as carotid intima-media thickness, coronary artery calcification, and cardiac troponin T. The absence of systolic blood pressure data precluded a direct comparison of our model with established risk scores such as the PCE, SCORE2, and SCORE2-OP, which incorporate blood pressure as a key variable [5, 13]. While the utility of family history in enhancing CV risk prediction has been questioned, the American Heart Association continues to recommend its consideration as a 'risk-enhancing factor' in clinical decision-making [2, 46]. The practical applicability of emerging biomarkers remains uncertain [7, 10], necessitating further research to determine the need to integrate them into predictive models with consideration of their feasibility in routine clinical practice. Future studies should also validate and compare our model with existing risk scores in populations with complete datasets that include all necessary variables.

Conclusion

In summary, this study provided a sex-specific five-year CV risk assessment tool specifically designed for Asian individuals aged 75 years and older. This tool exhibited fair discriminatory ability for both sexes, with particularly good calibration among women. The identification of significant risk factors suggests potential benefits for guiding targeted preventive interventions for individuals at elevated CV risk. To further enhance the tool’s robustness and applicability, external validation using an independent dataset is warranted.

Data availability

This study was based on the national claims database provided by the National Health Insurance Administration (NHIA) and the electronic health records from a medical center in Taiwan. Data cannot be made publicly available due to legal restrictions.

Abbreviations

CVD:

Cardiovascular disease

CV:

Cardiovascular

PCE:

Pooled Cohort Equations

SCORE:

Systematic Coronary Risk Evaluation

QRISK:

QRESEARCH Cardiovascular Risk Algorithm

NTUH-iMD:

National Taiwan University Hospital-integrative Medical Database

NHIRD:

National Health Insurance Research Database

NHI:

National Health Insurance

MI:

Myocardial infarction

ICD-9-CM:

International Classification of Diseases, Ninth Revision, Clinical Modification

ICD-9-PCS:

International Classification of Diseases, Ninth or Tenth Revision, Procedure Coding System

MACE:

Major adverse cardiovascular events

BMI:

Body mass index

LDL-C:

Low-density lipoprotein cholesterol

HDL-C:

High-density lipoprotein cholesterol

T-CHO:

Total cholesterol

TG:

Triglycerides

COPD:

Chronic obstructive pulmonary disease

mFI:

Multimorbidity Frailty Index

ATC:

Anatomical Therapeutic Chemical

SD:

Standard deviation

AIC:

Akaike Information Criterion

AUC:

Area under the receiver operating curve

C-index:

Harrell's concordance index

PY:

Person-years

References

  1. Ministry of Health and Welfare, Taiwan. Causes of death report. 2017. Available from: https://www.mohw.gov.tw/cp-16-41794-1.html.

  2. Grundy SM, Stone NJ, Bailey AL, Beam C, Birtcher KK, Blumenthal RS, et al. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA guideline on the management of blood cholesterol: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol. 2019;73(24):3168–209.

    Article  PubMed  Google Scholar 

  3. Visseren FLJ, Mach F, Smulders YM, Carballo D, Koskinas KC, Bäck M, et al. 2021 ESC guidelines on cardiovascular disease prevention in clinical practice. Eur Heart J. 2021;42(34):3227–337.

    Article  PubMed  Google Scholar 

  4. National Institute for Health and Care Excellence (NICE), UK. Cardiovascular disease: risk assessment and reduction, including lipid modification. NICE guideline. 2023. Available from: https://www.nice.org.uk/guidance/ng238.

  5. DeFilippis AP, Michos ED, McEvoy JW, Blaha MJ, Young R, Kronmal RA, et al. Risk score overestimation: the impact of individual cardiovascular risk factors and preventive therapies on the performance of the American Heart Association-American College of Cardiology-Atherosclerotic Cardiovascular Disease risk score in a modern multi-ethnic cohort. Eur Heart J. 2016;38(8):598–608.

    PubMed Central  Google Scholar 

  6. Veronese N, Cereda E, Stubbs B, Solmi M, Luchini C, Manzato E, et al. Risk of cardiovascular disease morbidity and mortality in frail and pre-frail older adults: Results from a meta-analysis and exploratory meta-regression analysis. Ageing Res Rev. 2017;35:63–73.

    Article  PubMed  PubMed Central  Google Scholar 

  7. Barry AR, O’Neill DE, Graham MM. Primary prevention of cardiovascular disease in older adults. Can J Cardiol. 2016;32(9):1074–81.

    Article  PubMed  Google Scholar 

  8. Poortvliet RK, de Ruijter W, de Craen AJ, Mooijaart SP, Westendorp RG, Assendelft WJ, et al. Blood pressure trends and mortality: the Leiden 85-plus Study. J Hypertens. 2013;31(1):63–70.

    Article  PubMed  CAS  Google Scholar 

  9. Bell SP, Saraf A. Risk stratification in very old adults: how to best gauge risk as the basis of management choices for patients aged over 80. Prog Cardiovasc Dis. 2014;57(2):197–203.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Saeed A, Nambi V, Sun W, Virani SS, Taffet GE, Deswal A, et al. Short-term global cardiovascular disease risk prediction in older adults. J Am Coll Cardiol. 2018;71(22):2527–36.

    Article  PubMed  PubMed Central  Google Scholar 

  11. Cooney MT, Selmer R, Lindman A, Tverdal A, Menotti A, Thomsen T, et al. Cardiovascular risk estimation in older persons: SCORE O.P. Eur J Prev Cardiol. 2015;23(10):1093–103.

    Article  PubMed  Google Scholar 

  12. Huynh QL, Reid CM, Chowdhury EK, Huq MM, Billah B, Wing LM, et al. Prediction of cardiovascular and all-cause mortality at 10 years in the hypertensive aged population. Am J Hypertens. 2015;28(5):649–56.

    Article  PubMed  Google Scholar 

  13. SCORE2-OP working group and ESC Cardiovascular risk collaboration. SCORE2-OP risk prediction algorithms: estimating incident cardiovascular event risk in older persons in four geographical risk regions. Eur Heart J. 2021;42(25):2455–67.

    Article  Google Scholar 

  14. DeFilippis AP, Young R, Carrubba CJ, McEvoy JW, Budoff MJ, Blumenthal RS, et al. An analysis of calibration and discrimination among multiple cardiovascular risk scores in a modern multiethnic cohort. Ann Intern Med. 2015;162(4):266–75.

    Article  PubMed  PubMed Central  Google Scholar 

  15. Choi J, Sung S, Park SK, Park S, Kim H, Cho M-C, et al. SCORE and SCORE2 in East Asian population: a performance comparison. JACC Asia. 2024;4(4):265–74.

    Article  PubMed  Google Scholar 

  16. Jung KJ, Jang Y, Oh DJ, Oh B-H, Lee SH, Park S-W, et al. The ACC/AHA 2013 pooled cohort equations compared to a Korean Risk Prediction Model for atherosclerotic cardiovascular disease. Atherosclerosis. 2015;242(1):367–75.

    Article  PubMed  CAS  Google Scholar 

  17. Yang X, Li J, Hu D, Chen J, Li Y, Huang J, et al. Predicting the 10-year risks of atherosclerotic cardiovascular disease in Chinese population: the China-PAR Project (Prediction for ASCVD Risk in China). Circulation. 2016;134(19):1430–40.

    Article  PubMed  Google Scholar 

  18. Wu Y, Liu X, Li X, Li Y, Zhao L, Chen Z, et al. Estimation of 10-year risk of fatal and nonfatal ischemic cardiovascular diseases in Chinese adults. Circulation. 2006;114(21):2217–25.

    Article  PubMed  Google Scholar 

  19. Pinho-Gomes AC, Peters SAE, Thomson B, Woodward M. Sex differences in prevalence, treatment and control of cardiovascular risk factors in England. Heart. 2021;107:462–7.

    Article  Google Scholar 

  20. Peters SAE, Muntner P, Woodward M. Sex differences in the prevalence of, and trends in, cardiovascular risk factors, treatment, and control in the United States, 2001 to 2016. Circulation. 2019;139(8):1025–35.

    Article  PubMed  CAS  Google Scholar 

  21. Appelman Y, van Rijn BB, Ten Haaf ME, Boersma E, Peters SA. Sex differences in cardiovascular risk factors and disease prevention. Atherosclerosis. 2015;241(1):211–8.

    Article  PubMed  CAS  Google Scholar 

  22. Betai D, Ahmed AS, Saxena P, Rashid H, Patel H, Shahzadi A, et al. Gender disparities in cardiovascular disease and their management: a review. Cureus. 2024;16(5):e59663.

    PubMed  PubMed Central  Google Scholar 

  23. Lee YC, Chao YT, Lin PJ, Yang YY, Yang YC, Chu CC, et al. Quality assurance of integrative big data for medical research within a multihospital system. J Formos Med Assoc. 2022;121(9):1728–38.

    Article  PubMed  Google Scholar 

  24. Brugts JJ, Yetgin T, Hoeks SE, Gotto AM, Shepherd J, Westendorp RGJ, et al. The benefits of statins in people without established cardiovascular disease but with cardiovascular risk factors: meta-analysis of randomised controlled trials. BMJ. 2009;338:b2376.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  25. Wen YC, Chen LK, Hsiao FY. Predicting mortality and hospitalization of older adults by the multimorbidity frailty index. PLoS ONE. 2017;12(11):e0187825.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Sterne JA, White IR, Carlin JB, Spratt M, Royston P, Kenward MG, et al. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ. 2009;338:b2393.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Lloyd-Jones DM. Cardiovascular risk prediction: basic concepts, current status, and future directions. Circulation. 2010;121(15):1768–77.

    Article  PubMed  Google Scholar 

  28. Vergouwe Y, Steyerberg EW. Towards better clinical prediction models: seven steps for development and an ABCD for validation. Eur Heart J. 2014;35(29):1925–31.

    Article  PubMed  PubMed Central  Google Scholar 

  29. Harrell FE Jr, Califf RM, Pryor DB, Lee KL, Rosati RA. Evaluating the yield of medical tests. JAMA. 1982;247(18):2543–3256.

    Article  PubMed  Google Scholar 

  30. Buttner S, Galjart B, Beumer BR, van Vugt JLA, van Eijck CHJ, Polak WG, et al. Quality and performance of validated prognostic models for survival after resection of intrahepatic cholangiocarcinoma: a systematic review and meta-analysis. HPB (Oxford). 2021;23(1):25–36.

    Article  PubMed  Google Scholar 

  31. Cook NR. Use and misuse of the receiver operating characteristic curve in risk prediction. Circulation. 2007;115(7):928–35.

    Article  PubMed  Google Scholar 

  32. Crowson CS, Atkinson EJ, Therneau TM. Assessing calibration of prognostic risk scores. Stat Methods Med Res. 2016;25(4):1692–706.

    Article  PubMed  Google Scholar 

  33. Zhang Y, Miao H, Chia YC, Buranakitjaroen P, Siddique S, Shin J, et al. Cardiovascular risk assessment tools in Asia. Journal of Clinical Hypertension (Greenwich). 2022;24(4):369–77.

    Article  Google Scholar 

  34. Zhiting G, Jiaying T, Haiying H, Yuping Z, Qunfei Y, Jingfen J. Cardiovascular disease risk prediction models in the Chinese population- a systematic review and meta-analysis. BMC Public Health. 2022;22(1):1608.

    Article  PubMed  PubMed Central  Google Scholar 

  35. Zhao D. Epidemiological features of cardiovascular disease in Asia. JACC Asia. 2021;1(1):1–13.

    Article  PubMed  PubMed Central  Google Scholar 

  36. Boekholdt SM, Arsenault BJ, Mora S, Pedersen TR, LaRosa JC, Nestel PJ, et al. Association of LDL cholesterol, non-HDL cholesterol, and apolipoprotein B levels with risk of cardiovascular events among patients treated with statins: a meta-analysis. JAMA. 2012;307(12):1302–9.

    Article  PubMed  CAS  Google Scholar 

  37. Lin FJ, Tseng WK, Yin WH, Yeh HI, Chen JW, Wu CC. Residual Risk Factors to Predict Major Adverse Cardiovascular Events in Atherosclerotic Cardiovascular Disease Patients with and without Diabetes Mellitus. Sci Rep. 2017;7(1):9179.

    Article  PubMed  PubMed Central  Google Scholar 

  38. Lind L, Sundström J, Ärnlöv J, Lampa E. Impact of aging on the strength of cardiovascular risk factors: a longitudinal study over 40 years. J Am Heart Assoc. 2018;7(1):e007061.

    Article  PubMed  PubMed Central  Google Scholar 

  39. Sachdeva A, Cannon CP, Deedwania PC, LaBresh KA, Smith SC Jr, Dai D, et al. Lipid levels in patients hospitalized with coronary artery disease: an analysis of 136,905 hospitalizations in Get With The Guidelines. Am Heart J. 2009;157(1):111-7. e2.

    Article  PubMed  CAS  Google Scholar 

  40. Thompson W, Pottegard A, Nielsen JB, Haastrup P, Jarbol DE. How common is statin use in the oldest old? Drugs Aging. 2018;35(8):679–86.

    Article  PubMed  CAS  Google Scholar 

  41. Hippisley-Cox J, Coupland C, Vinogradova Y, Robson J, Minhas R, Sheikh A, et al. Predicting cardiovascular risk in England and Wales: prospective derivation and validation of QRISK2. BMJ. 2008;336(7659):1475–82.

    Article  PubMed  PubMed Central  Google Scholar 

  42. van Bussel EF, Richard E, Busschers WB, Steyerberg EW, van Gool WA, Moll van Charante EP, et al. A cardiovascular risk prediction model for older people: Development and validation in a primary care population. J Clin Hypertens. 2019;21(8):1145–52.

    Article  Google Scholar 

  43. Roger VL, Go AS, Lloyd-Jones DM, Adams RJ, Berry JD, Brown TM, et al. Heart disease and stroke statistics–2011 update: a report from the American Heart Association. Circulation. 2011;123(4):e18–209.

    PubMed  Google Scholar 

  44. Maas AH, Franke HR. Women’s health in menopause with a focus on hypertension. Neth Hear J. 2009;17(2):68–72.

    Article  CAS  Google Scholar 

  45. Engberding N, Wenger NK. Management of hypertension in women. Hypertens Res. 2012;35(3):251–60.

    Article  PubMed  Google Scholar 

  46. Sivapalaratnam S, Boekholdt SM, Trip MD, Sandhu MS, Luben R, Kastelein JJ, et al. Family history of premature coronary heart disease and risk prediction in the EPIC-Norfolk prospective population study. Heart. 2010;96(24):1985–9.

    Article  PubMed  Google Scholar 

Download references

Acknowledgements

This study is based, in part, on data from the National Health Insurance Research Database provided by the National Health Insurance Administration, Taiwan’s Ministry of Health, and Welfare and managed by the Health and Welfare Data Science Center. The interpretation and conclusions contained herein do not represent those of the National Health Insurance Administration or the Health and Welfare Data Science Center. The authors would like to express their appreciation to the staff at the Department of Medical Research for providing clinical data from the National Taiwan University Hospital-integrative Medical Database. Additionally, they would like to thank Chi-Ying Wu for his assistance in revising the manuscript.

Funding

The work was partially supported by Taiwan’s Ministry of Science and Technology [grant number 106–2320-B-002–032-MY3]. The funders had no role in the study design, data collection, analysis, result interpretation, publication decision, or manuscript preparation.

Author information

Authors and Affiliations

  1. Graduate Institute of Clinical Pharmacy, College of Medicine, National Taiwan University, No. 33, Linsen S. Rd., Zhongzheng Dist., Taipei, 100025, Taiwan

    Meng-Chen Hsu, Yu-Hua Fu, Chi-Chuan Wang & Fang-Ju Lin

  2. School of Pharmacy, College of Medicine, National Taiwan University, Taipei, Taiwan

    Chi-Chuan Wang & Fang-Ju Lin

  3. Department of Pharmacy, National Taiwan University Hospital, Taipei, Taiwan

    Chi-Chuan Wang & Fang-Ju Lin

  4. Division of Cardiology, Department of Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan

    Chau-Chung Wu

  5. Department/Graduate Institute of Medical Education and Bioethics, College of Medicine, National Taiwan University, Taipei, Taiwan

    Chau-Chung Wu

Authors
  1. Meng-Chen Hsu
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Yu-Hua Fu
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Chi-Chuan Wang
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Chau-Chung Wu
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Fang-Ju Lin
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

M.-C. H. and F.-J. L. had full access to all the data in the study and took responsibility for the integrity and accuracy of the data analysis. F.-J. L. supervised the conduct of the study. M.-C. H. designed the study, carried out the data analysis, and drafted the initial manuscript. All the authors contributed to the study design and data interpretation and provided critical review and revision of the manuscript for intellectual content. All the authors approved the submission of the final manuscript.

Corresponding author

Correspondence to Fang-Ju Lin.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

This study was presented in part at the 12th Asian-Pacific Society of Atherosclerosis and Vascular Disease (APSAVD) Congress, Taipei, Taiwan, September 20-22, 2019.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Hsu, MC., Fu, YH., Wang, CC. et al. Development and validation of a five-year cardiovascular risk assessment tool for Asian adults aged 75 years and older. BMC Geriatr 25, 15 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12877-024-05660-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12877-024-05660-4

Keywords

BMC Geriatrics

ISSN: 1471-2318

Contact us