LABORATORIO DE URGENCIAS

Prognostic Value of Cardiovascular Biomarkers in the Population

                                                                    Johannes Tobias Neumann, MD, PhD; Raphael Twerenbold, MD; JessicaWeimann, MSc;
                                                           Christie M. Ballantyne, MD; Emelia J. Benjamin, MD, Francisco Ojeda, PhD et al

                                                                                                 JAMA Published online May, 2024

Early identification of individuals in the general population at high risk for atherosclerotic cardiovascular disease shapes primary preventive strategies to reduce the risk ofdevelopingatherosclerotic cardiovascular disease.1,2 Risk scores based on traditional risk factors for atherosclerotic cardiovascular disease (eg, the European Society of Cardiology Systematic CoronaryRisk Evaluation2[SCORE2],theAmerican Heart Association/American College of Cardiology Pooled Cohort Equations, and the American Heart Association Predicting Risk of Cardiovascular Disease Events [PREVENT] equations) are widely available to estimate an individual’s risk for future cardiovascular events.1-4
Cardiovascular biomarkers, such as cardiac troponin, natriuretic peptides, and C reactive protein (CRP), are established in clinical care. Using newer, high-sensitivity cardiac troponin assays, concentrations became measurable in the general population, opening up the prospects for a broader application of this biomarker.5 Several studies have reported strong associations of these biomarkers with incident atherosclerotic cardiovascular disease events in individuals with known atherosclerotic cardiovascular disease, but also, and most importantly, in apparently healthy individuals and an improvement in risk stratification when these biomarkers were added to established risk prediction models.5-12 Notwithstanding the achievements of earlier studies, the actual application of routinely available cardiovascular biomarkers for risk stratification in primary prevention has not become routine clinical practice. In addition, it remains uncertain which of the established biomarkers might be best suited to predict each outcome and howsuch associations are influenced by age.
Therefore, this study brings together the largest multinational individual-level dataset, to date, to investigate the comparative predictive value of cardiovascular biomarkers for incident atherosclerotic cardiovascular disease events in the general population and to elucidate their differential effects according to age.

METHODS

*Study Cohorts
This individual-level analysis included data from 28 multinational population-based cohorts (eTable 1 in Supplement 1). Eligible cohorts were those that included:
(1) individuals from the general population, in which most participants were apparently healthy (ie, had not had any major atherothrombotic cardiovascular events).
(2) individuals who had at least 1 measurement of high-sensitivity cardiac troponin I, high-sensitivity cardiac troponin T, B-type natriuretic peptide (BNP), N-terminal pro-BNP (NT-proBNP), or high-sensitivity CRP.
(3) individuals with follow-up for at least 2 years. Data from all cohorts were collected and harmonized in a database. Individuals with a history of atherosclerotic cardiovascular disease events or heart failure were excluded from the analyses

*Study Outcomes
The primary outcome was incident atherosclerotic cardiovascular disease, which included all fatal and nonfatal events. Incident atherosclerotic cardiovascular disease was defined by the first possible or definite coronary heart disease event, possible or definite ischemic stroke event, coronary revascularization, coronary heart disease death, ischemic stroke death, or unclassifiable death.13
The secondary outcomes were all-cause mortality, incident heart failure, incident ischemic stroke, and incident myocardial infarction. Additional information about the outcomes investigated in Supplement 1.

*Biomarkers
For all cohorts reported by the Biomarker for Cardiovascular Risk Assessment across Europe (BiomarCaRE) consortium, serum high-sensitivity cardiac troponin I concentration was determined in the BiomarCaRE core laboratory in Hamburg, Germany, using a highly sensitive cardiac troponin I immunoassay for stored samples (Architect i2000SR; Abbott Diagnostics). The limit of detection for the immunoassay was 1.9 ng/L (range, 0-50 000 ng/L) and the assay had a coefficient of variation of 10% at a concentration of 5.2 ng/L.
Measurement of NT-proBNP concentration was performed using an electrochemiluminescence immunoassay (ELECSYS 2010 and Cobas e411; Roche Diagnostics); the analytic range is 0 ng/L to 35000 ng/L.
Measurement of high-sensitivity CRP concentration was performed using the Vario immunoassay and the Architect c8000 system (Abbott Diagnostics).
For the cohorts not part of the BiomarCaRE consortium, measurements of highsensitivity cardiac troponin I, high-sensitivity cardiac troponin T, NT-proBNP, BNP, and high-sensitivity CRP were performed as part of local cohort–specific procedures (details in Supplement 1).

*Statistical Analyses
Associations Between Cardiac-Specific Biomarkers and Study Outcomes
To examine the unadjusted association of the biomarkers and the primary outcome, cumulative incidence curveswere computed according to biomarker quintiles. Death from causes unrelatedto atherosclerotic cardiovascular disease was treated as a competing event. The curves were estimated using the Aalen-Johansen estimator.
The quintileswere computed using linear quantile mixed models.14,15 Fine and Gray subdistribution hazard models were calculated.
Death from causes unrelated to atherosclerotic cardiovascular disease or the secondary outcomes were treated as a competing event, respectively.16
Cox proportional hazards regression models were used for all-cause mortality.
The biomarkers were used as continuous variables after applying loge transformation with hazard ratios (HRs) or subdistribution HRs computed additionally per 1-SD change to allow for comparisons of the effect size among different biomarkers.
The models with high-sensitivity cardiac troponin concentration as a continuous variable were augmented with a binary variable indicating if the measured value was above or below the limit of detection.17
The regression models for all the study outcomeswere adjusted for sex and cohort as stratification variables. The models were also adjusted for age, total cholesterol, high-density lipoprotein (HDL) cholesterol, current smoking, prevalent diabetes, systolic blood pressure, and self-reported use of antihypertensive drugs. For the outcomes of all-cause mortality and heart failure, the models were additionally adjusted for body mass index (calculated asweight in kilograms divided by height in meters squared).
In separate analyses, high-sensitivity cardiac troponin I, NT-proBNP and highsensitivity CRP were combined in a multivariable model because these biomarkers represent different pathophysiological pathways and the data were readily available in the cohorts. The time-to-event modelswere extended by modeling the biomarkers using penalized cubic splines.

*Added Predictive Value
The C statistic and net reclassification improvement (NRI) were used to quantify the added predictive value of each biomarker beyond that from the adjusted model described above. In the presence of competing risks, the calculations of the C statistic and NRI were adapted.18,19 Internal-external cross-validation was applied to control for the overoptimism of calculating performance measures on the same dataset from which the models were computed.20 Namely, each study was in turn left out and the Cox model or the Fine and Gray model was estimated in the remaining studies. Next, the models were used to estimate the event probabilities in the excluded study.
The category-based NRI and the continuous NRI were calculated. The 95% CIs for the C statistic and NRI were computed by bootstrapping 500 times the internal external cross-validation.
All statistical analyses were performed using R versión 4.2.2 (R Foundation for Statistical Computing).21 Additional information about the statistical analyses appears in Supplement 1.

RESULTS

*Study Population
There were 28 general population–based cohorts included (from 12 countries and 4 continents) with data on 164 054 individuals (median age, 53.1 years [IQR, 42.7-62.9 years]; 85 972 [52.4%] were women; and 9977 [6.1%] had diabetes) (more details in Supplement 1).
Of the 162 947 individuals with data for the hypertension variable, 67 719 (41.6%) had hypertension. Of the 162 139 individuals with data for the smoking variable, 40 226 (24.8%) smoked daily. The median 10-year atherosclerotic cardiovascular disease risk SCORE2 was 4.1% (IQR, 1.7%-8.6%) and the corresponding 10-year risk using the Pooled Cohort Risk Equation was 4.9% (IQR, 1.4%-13.1%).
The median biomarker concentrations were 2.5 ng/L (IQR, 1.9-4.1 ng/L) for high-sensitivity cardiac troponin I, 3.1 ng/L (IQR, 3.0-6.0 ng/L) for high-sensitivity cardiac troponin T, 43.8 ng/L (IQR, 20.6-86.2 ng/L) for NT-proBNP,14.9 ng/L (IQR, 7.9-28.6 ng/L) for BNP, and 1.4mg/L (IQR, 0.7-3.2 mg/L) for high-sensitivity CRP ( more details  in Supplement 1).
Except for NT-proBNP, there was a linear relationship between age and the median biomarker concentrations (Table in Supplement 1).
During a median follow-up of 11.8 years (IQR, 6.2-18.0 years; maximum follow-up, 28.2 years), there were 17 211 incident atherosclerotic cardiovascular disease events, 25 346 deaths from any cause, 6766 cases of heart failure, 4794 incident cases of incident ischemic stroke,and 8024 incident cases of myocardial infarction (Table 5 in Supplement 1).

*Primary Outcome: Association of Biomarkers With Incident Atherosclerotic Cardiovascular Disease Events
After adjusting for sex and cohort and the conventional risk factors of age, total cholesterol,HDLcholesterol, smoking status, diabetes, systolic blood pressure, and self-reported use of antihypertensive drugs, the biomarker concentrations were associated with incident atherosclerotic cardiovascular disease events (subdistributionHRper 1-SDchange, 1.13 [95%CI, 1.11-1.16] for high-sensitivity cardiac troponin I; 1.18 [95%CI, 1.12-1.23] for high-sensitivity cardiac troponin T; 1.21 [95% CI,1.18-1.24] for NT proBNP; 1.14 [95% CI, 1.08-1.22] for BNP; and 1.14 [95% CI, 1.12-1.16] for high-sensitivity CRP
For all 5 biomarkers, therewere more events per 1000 person-years in individuals with biomarker concentrations above the median comparedwith thosewith biomarker concentrations below the median. Additional data in eTable 6 in Supplement 1.
In separate analyses, high-sensitivity cardiac troponin I, NT-proBNP,andhigh-sensitivityCRPwere included in thesame model and proved to be predictors of atherosclerotic cardiovascular disease events (adjusted subdistribution HR, 1.07 [95%CI, 1.04-1.10] for high-sensitivity cardiac troponin I; 1.19 [95%CI, 1.15-1.23] for NT-proBNP; and 1.14 [95%CI, 1.10-1.17] for high-sensitivity CRP). When stratified by biomarker quintiles, the cumulative atherosclerotic cardiovascular disease incidence gradually increasedwith increasing biomarker concentrations (in Supplement 1).
The addition of the biomarkers to the base model, which included only conventional risk factors, was associated with an increase in the C statistics for atherosclerotic cardiovascular disease events after 1 year, 5 years, and 10 years (Table 7 in Supplement 1).
The strongest increase was observed when high-sensitivity cardiac troponin I, NT-proBNP, and high-sensitivity CRP were combined in 1 model. In the reclassification analyses, the categorical NRI for the combination of high-sensitivity cardiac troponin I, NT-proBNP and high-sensitivityCRPwas0.044(95%CI,0.023-0.069) (in Supplement 1).
The continuous NRI was 0.241 (95% CI, 0.193-0.309) for high-sensitivity cardiac troponin I, 0.201 (95% CI, 0.008- 0.364) for high-sensitivity cardiac troponin T, 0.06 (95% CI, 0.016-0.093) for NT-proBNP, 0.077 (95% CI, 0.000-0.144) for BNP, 0.192 (95% CI, 0.162-0.222) for high-sensitivity CRP, and 0.23 (95% CI, 0.162-0.283) for the combination of high-sensitivity cardiac troponin I, NT-proBNP, and highsensitivity CRP (in Supplement 1).

*Association of Biomarkers With Secondary Outcomes
All biomarkers were associated with all-cause mortality, incident heart failure, incident ischemic stroke, and incident myocardial infarction (in Supplement 1).
The associations of biomarkers with all-cause mortality and incident heart failure were larger than those for atherosclerotic cardiovascular disease. The highest subdistribution HRs were observed for incident heart failure with high-sensitivity cardiac troponin T (HR, 1.44; 95% CI, 1.38-1.51), NT-proBNP (HR,1.62; 95%CI, 1.56-1.68), and BNP (HR, 1.59; 95%CI, 1.43-1.77).
The addition of the biomarkers also improved the C statistics and the appropriate classification of risk in Supplement 1)when added to the base model for the secondary outcomes.
The largest risk classification improvements were for heart failure and all-cause mortality.

*Sensitivity Analyses
When stratified according to the age cutoff of65 years, older individuals (n = 34 143; aged ≥65 years) more often had diabetes and hypertension, but smoked less often than younger individuals (n = 129 456; aged <65 years) (in Supplement1). Concentrations of high-sensitivity cardiac troponin I, high-sensitivity cardiac troponinT ,NT-proBNP,BNP, highsensitivity CRP were higher, on average, in older individuals.
The association of biomarkers with atherosclerotic cardiovascular disease events remained significant in both individuals younger than 65 years of age and in those aged 65 years or older (in Supplement 1). The subdistribution HRs for high-sensitivity cardiac troponin I, highsensitivity cardiac troponin T, and NT-proBNP were higher in older individuals. The subdistribution HR was lower for highsensitivity CRP in older people. In older people, the C statistic of the base model was substantially lower compared with younger people and the addition of the biomarkers provided higher absolute increases of the C statistic in older people (in Supplement 1). For example, the combination of high-sensitivity cardiac troponin I, NT-proBNP, and highsensitivity
CRP increased the C statistic for 10-year atherosclerotic cardiovascular disease events from 0.812 (95% CI, 0.8021-0.8208) to 0.8194 (95% CI, 0.8089-0.8277) in younger people and from 0.6323 (95% CI, 0.5945-0.6570) to 0.6602 (95% CI, 0.6224-0.6834) in older people. The overall NRI with the combined biomarkers for atherosclerotic cardiovascular disease was higher in older people (NRI, 0.062; 95% CI, 0.013-0.120) compared with younger people (NRI, 0.028; 95% CI, 0.010-0.070) (in Supplement 1).
The HRs for all-cause mortality were also higher for NT-proBNP and BNP in older people, but slightly lower for high-sensitivity CRP and high-sensitivity cardiac troponinT (in Supplement 1). A similar pattern was also observed for the other outcomes of heart failure, ischemic stroke, and myocardial infarction. The overall NRI using the combination of high-sensitivity cardiac troponin I, NT-proBNP, and high-sensitivity CRP was higher in older people for the secondary outcomes of heart failure, ischemic stroke, andmyocardial infarction, but lower for all-cause mortality (in Supplement 1).
In sensitivity analyses, the regression analyses were repeatedfor 10-year atherosclerotic cardiovascular disease events while removing 1 risk factor from the base model (in Supplement 1). The addition of high-sensitivity cardiac troponin I, NT-proBNP, and high-sensitivity CRP resulted in the highest increase in the C statistic when age was removed from the base model (C statistic difference of 0.0312; 95%CI).
Sensitivity analyses also were performed that included information on cholesterol-lowering medication, which was available in 71.2%of the study population (in Supplement1). These findings were consistent with the primary study findings.

DISCUSSION

In this individual-level analysis, the value of the most commonly used biomarkers for cardiovascular risk prediction in the general population was investigated using harmonized, multinational population data (from 28 cohorts in 12 countries and 4 continents).
There were 4 salient findings:
* First, all investigated biomarkers were predictors not only of incident atherosclerotic cardiovascular disease events, but also of all-cause mortality, heart failure,myocardial infarction, and ischemic stroke. Even though prior studies from the general population focused on the association of biomarkers with fatal or nonfatal atherosclerotic cardiovascular disease in general, most did not consider other important outcomes. 7,8,22,23
Previous analyses examining incident heart failure or ischemic strokewere limited by:
(1) small numbers of events.
(2) the availability of aggregate data only
(3) shorter duration of follow-up.7
Interestingly, there was a stronger association of all the investigated biomarkers with all-cause mortality,and particularly with heart failure, comparedwith fatal and nonfatal atherosclerotic cardiovascular disease events.
In the current data set, all-causemortalitywasthemostfrequently reported outcome(25 346events), highlighting thepotential competing risk of death for any regression analyses.
Thus, the consideration of competing risk by using Fine and Gray regression analyses may be a possible explanation for opposite results compared with prior studies10,22,24 and strengthens the findings fromthe current analysis. The strong predictive value for heart failure outcomes is particularly notable given the increase in options available for preventing incident heart failure (such as intensive blood pressure control and treatment with sodium-glucose cotransporter 2 inhibitors).25
Importantly, the magnitude of change in the C statistic for the outcomes of all-cause mortality and heart failure reported in the current study is similar to other studies that investigated the addition of coronary calcium scoring to classic cardiovascular risk factors to predict atherosclerotic cardiovascular disease.26
*Second significant finding was that the combination of the biomarkers high-sensitivity cardiac troponin I,NT-proBNP, and high-sensitivity CRP into 1 model provided the largest incremental predictive value and that all 3 biomarkers were independent predictors. These 3 biomarkers represent 3 different pathophysiological pathways, had the highest availability in the cohorts examined, are routinely available, andwere also
identified as the strongest predictors in earlier analyses of multiple biomarkers.27.Most prior studies focused on 1 specific cardiovascular biomarker and did not attempt to combine several markers into 1model.28
The combination of high-sensitivity cardiac troponin I,NT-proBNP, and high-sensitivity CRP in the current study resulted in the biggest improvements in the C statistic for most outcomes investigated.
Interestingly, the multivariable model showed the highest HRs for NT-proBNP for all outcomes except for incident myocardial infarction for which high-sensitivity CRP showed the strongest association. This ranking of biomarkers is comparable with prior analyses from the FINRISK and Belfast PRIME cohorts,27 for which the highest HRs for atherosclerotic cardiovascular disease events were observed for NTproBNP; however,nohigh-sensitivity troponin assay was available for comparison at that time.
*Third important finding is a sustained association of improved risk prediction when biomarkers were added to conventional risk factors over a time horizon of more than 10 years.
The long follow-up (median duration of nearly 12 years) enabled the assessment of the C statistic over a long time frame, and the incremental value of the biomarkerswasapparent even beyond10years. This observation highlights the potential value of biomarkers for incorporation in primary prevention strategies,
Which ideally shouldaddress long-termeffects.Prior post hoc analyses from the JUPITER,WOSCOPS, and SPRINT large clinical trials29-31 investigated the role of biomarkers, especially cardiac troponin and NT-proBNP for decision-making in preventive care. Data from the JUPITER trial29 showed that those individuals with higher concentrations of highsensitivity cardiac troponin I or BNP were at higher risk for atherosclerotic cardiovascular disease events and may have a higher absolute risk reduction with statin treatment.
In theWOSCOPS trial,30 longitudinal measurementswere available for high-sensitivity cardiac troponin I and showed an association with atherosclerotic cardiovascular disease events and also their decrease 1 year after statin treatment. Recently,post hoc analyses from the SPRINT trial revealed that individuals with elevated concentrations of high-sensitivity cardiac troponin I and NT-proBNP had a substantially increased risk of all-cause mortality and heart failure, but also had the highest absolute risk reductionwith treatment comparedwith individuals with normal concentrations of the biomarkers.31
*Fourth novel finding is the greater incremental value of biomarkers in older individuals (aged ≥65 years) compared with younger individuals (aged<65years).Prior studies showed that with increasing age, the effect of conventional risk factors is attenuated. 3,32,33 In the current study, the conventional risk factor model had a C statistic of 0.632 in older individuals vs 0.812 in younger individuals. This resulted in the development of risk prediction models specifically for older people.3,34
There remains substantial residual risk when predicting incident atherosclerotic cardiovascular disease events, high lighting the need for other clinically relevant risk markers
In this context, the current study findings support the relevance of cardiovascular biomarkers, especially in older individuals.
This observation was primarily drivenby the increasing predictive value of NT-proBNP in older people, whereas the predictive value of high-sensitivity CRP decreased. Importantly, these findings were not limited to atherosclerotic cardiovascular disease events, but were also observed for all of the secondary outcomes, especially all-cause mortality and heart failure.

LIMITATIONS

This analysis has limitations.
First, this study used 5 established biomarkers that are widely available in routine clinical practice; however, the absolute measurements for highsensitivity cardiac troponin T and BNP were limited.
Second, most individuals recruited from high-income cohorts in Europe and North America, which limits the worldwide generalizability of the findings.
Third, there were a limited number of Black participants. Non-Black participants systematically have higher concentrations of high-sensitivity CRP and higher absolute risks.
Fourth, important questions remain before cardiovascular biomarkers may be considered for implementation into clinical practice. These questions include the need for cost-effectiveness analyses and the identification of a target population.

CONCLUSIONS

Cardiovascular biomarkers were strongly associated with fatal and non fatal cardiovascular events and mortality.
The addition of biomarkers to established risk factors led to only a small improvement in risk prediction metrics for atherosclerotic cardiovascular disease, but was more favorable for heart failure and mortality.

NOTE: tables, graphs, supplement in the magazine mentioned at the beginning

REFERENCES

1. Visseren FLJ, Mach F, Smulders YM, et al; ESC National Cardiac Societies; ESC Scientific Document Group. 2021 ESC guidelines on cardiovascular disease prevention in clinical practice. Eur Heart J. 2021;42(34):3227-3337. doi:1093/eurheartj/ ehab484
2. Arnett DK, Blumenthal RS, AlbertMA, et al. 2019 ACC/AHA guideline on the primary prevention of cardiovascular disease: a report of the American College of Cardiology/ American Heart Association task force on clinical practice guidelines. Circulation.2019;140(11):e
596-e646. doi:1161/CIR. 0000000000000678
3. SCORE2Working Group and ESC Cardiovascular Risk Collaboration. SCORE2 risk prediction algorithms: new models to estimate 10-year risk of cardiovascular disease in Europe. Eur Heart J. 2021; 42(25):2439-2454. doi:1093/eurheartj/ehab309
4. Khan SS, Matsushita K, Sang Y, et al; Chronic Kidney Disease Prognosis Consortium and the American Heart Association Cardiovascular-Kidney- Metabolic Science Advisory Group. Development and validation of the American Heart Association’s PREVENT equations. Circulation. 2024;149(6):430-449. doi:1161/CIRCULATIONAHA.123.067626
5. Westermann D, Neumann JT, Sörensen NA, Blankenberg S. High-sensitivity assays for troponin in patients with cardiac disease. Nat Rev Cardiol. 2017;14(8):472-483. doi:1038/nrcardio.2017.48
6. Neumann JT, Twerenbold R, Ojeda F, et al; COMPASS-MI Study Group. Application of high-sensitivity troponin in suspected myocardial infarction. N Engl J Med. 2019;380(26):2529-2540. doi:1056/NEJMoa1803377
7. Willeit P,Welsh P, Evans JDW, et al. High-sensitivity cardiac troponin concentration and risk of first-ever cardiovascular outcomes in 154,052 participants. J AmColl Cardiol. 2017;70(5): 558-568. doi:10.1016/j.jacc.2017.05.062
8. Zeller T, Tunstall-Pedoe H, Saarela O, et al; MORGAM Investigators. High population prevalence of cardiac troponin I measured by a high-sensitivity assay and cardiovascular risk.The MORGAM Biomarker Project Scottish Cohort. Eur Heart J. 2014;35(5):271-281. doi:10.1093/eurheartj/eht406
9. Everett BM, Zeller T, Glynn RJ, Ridker PM, Blankenberg S. High-sensitivity cardiac troponin I and B-type natriuretic peptide as predictors of Prognostic Value of Cardiovascular Biomarkers in the PopulationJAMA Published online May 13, 2024 E11
10. Blankenberg S, Salomaa V, Makarova N, et al; BiomarCaRE Investigators. Troponin I and cardiovascular risk prediction in the general population: the BiomarCaRE consortium. Eur Heart J. 2016;37(30):2428-2437. doi:1093/eurheartj/ ehw172
11. Ridker PM, Buring JE, Rifai N, Cook NR. Development and validation of improved algorithms for the assessment of global cardiovascular risk in women: the Reynolds Risk Score. JAMA. 2007;297(6):611-619. doi:10.1001/ jama.297.6.611
12. Ridker PM, Paynter NP, Rifai N, Gaziano JM, Cook NR. C-reactive protein and parental history improve global cardiovascular risk prediction: the Reynolds Risk Score for men. Circulation. 2008;118 (22):2243-2251. doi:1161/CIRCULATIONAHA.108. 814251
13. National Institute for Health andWelfare and the MORGAM Project investigators. MORGAM manual. Accessed April 17, 2024. https://www.thl.fi/ publications/morgam/manual/contents.htm
14. Geraci M, Bottai M. Linear quantile mixed models. Stat Comput. 2014;24:461-479. doi: 1007/s11222-013-9381-9
15. Geraci M. Linear quantile mixed models: the lqmm package for Laplace quantile regression. J Stat Softw. 2014;57:1-29. doi:18637/jss.v057.i13
16. Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999 ;94(446):496-509. doi:10.1080/ 01621459.1999.10474144
17. Royston P, SauerbreiW. Multivariable Model-Building: A Pragmatic Approach to Regression Analysis Based on Fractional Polynomials for Modelling Continuous Variables. John Wiley & Sons; 2008.
18. Wolbers M, Koller MT, Witteman JCM, Steyerberg EW. Prognostic models with competing risks: methods and application to coronary risk prediction. Epidemiology. 2009;20(4):555-561.doi:1097/EDE.0b013e3181a39056
19. Pencina MJ, D’Agostino RB Sr, Steyerberg EW. Extensions of net reclassification improvement calculations to measure usefulness of new biomarkers.Stat Med. 2011; 30(1):11-21. doi:1002/sim.4085
20. Royston P, Parmar MKB, Sylvester R. Construction and validation of a prognostic model across several studies, with an application in superficial bladder cancer. Stat Med. 2004;23(6): 907-926. doi:1002/sim.1691
21. R Project for Statistical Computing. Getting started, news, news via Mastodon. Accessed April 17, 2024. http://www.R-project.org
22. Omland T, de Lemos JA, Holmen OL, et al. Impact of sex on the prognostic value of high-sensitivity cardiac troponin I in the general population: the HUNT study. Clin Chem. 2015;61(4): 646-656. doi:1373/clinchem.2014.234369
23. Welsh P, Preiss D, Shah ASV, et al. Comparison between high-sensitivity cardiac troponin T and cardiac troponin I in a large general population cohort. Clin Chem. 2018; 64(11):1607-1616. doi:1373/clinchem.2018.292086
24. de Lemos JA, Ayers CR, Levine BD, et al. Multimodality strategy for cardiovascular risk assessment: performance in 2 population-based cohorts. Circulation. 2017;135(22) :2119-2132. doi:1161/CIRCULATIONAHA.117.027272
25. McDonagh TA,Metra M, Adamo M, et al; ESC Scientific Document Group. 2021 ESC guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J. 2021;42(36):3599-3726.doi:1093/eurheartj/ehab368
26. Lin JS, Evans CV, Johnson E, Redmond N,Coppola EL, Smith N. Nontraditional risk factors in cardiovascular disease risk assessment: updated evidence report and systematic review for the US Preventive Services Task Force. JAMA. 2018;320(3): 281-297. doi:1001/jama.2018.4242
27. Blankenberg S, Zeller T, Saarela O, et al.Contribution of 30 biomarkers to 10-year cardiovascular risk estimation in 2 population cohorts: the MONICA, Risk, Genetics, Archiving,and Monograph (MORGAM) Biomarker Project.Circulation. 2010;121(22): 2388-2397. doi:10.1161/CIRCULATIONAHA.109.901413
28. Haller PM, Beer BN, Tonkin AM, Blankenberg S,Neumann JT. Role of cardiac biomarkers in epidemiology and risk outcomes. Clin Chem. 2021;67(1):96-106. doi:1093/clinchem/hvaa228
29. Everett BM, Zeller T, Glynn RJ, Ridker PM,Blankenberg S. High-sensitivity cardiac troponin I and B-type natriuretic peptide as predictors of vascular events in primary prevention: impact of statin therapy. Circulation. 2015;131(21):1851-1860. doi:10.1161/CIRCULATIONAHA.114.014522
30. Ford I, Shah ASV, Zhang R, et al. High-sensitivity cardiac troponin, statin therapy, and risk of coronary heart disease. J AmColl Cardiol. 2016;68(25):2719-2728. doi:10.1016/j.jacc.2016.10.020
31. Berry JD, Nambi V, AmbrosiusWT, et al.Associations of high-sensitivity troponin and natriuretic peptide levels with outcomes after intensive blood pressure lowering: findings from the SPRINT randomized clinical trial. JAMA Cardiol. 2021;6(12):1397-1405. doi:1001/jamacardio.2021.3187
32. Kannel WB, D’Agostino RB. The importance of cardiovascular risk factors in the elderly. Am J Geriatr Cardiol. 1995;4(2):10-23.
33. Lind L, Sundström J, Ärnlöv J, Lampa E. Impactof aging on the strength of cardiovascular risk factors: a longitudinal study over 40 years. J Am Heart Assoc. 2018;7(1):e007061. doi:1161/JAHA.117.00 7061
34. Neumann JT, Thao LTP, Callander E, et al. Cardiovascular risk prediction in healthy older people.Geroscience.2022;44(1):403-413. doi:1007/s11357-021-0048