ORIGINAL ARTICLE
Cardiovascular Risk Assessment in Primary Prevention in Patients with Type 2 Diabetes: Comparison of Multiple Scores and Detection of Subclinical Atheromatosis

Evaluación del riesgo cardiovascular en prevención primaria en pacientes con diabetes tipo 2: comparación de múltiples puntajes y detección de ateromatosis subclínica

  • LORENZO M. LOBO, 1  2  3  4  MTSAC,FACC, FESC  ORCID logo 
  • WALTER MASSON, 2  4  5  ORCID logo 
  • GRACIELA B. MOLINERO, 2  MTSAC  ORCID logo 
  • GUSTAVO GIUNTA, 1  4  MTSAC  ORCID logo 
  • AUGUSTO LAVALLE COBO, 2  4  6  ORCID logo 
  • CECILIA MASIELLO, 2  7 
  • JULIA D. DELGADO RIVEROS, 2 
  • VALENTINA SARLI, 2 
  • SILVINA SAGARDÍA, 2  MTSAC 
  • MELINA HUERIN, 2  4  9  MTSAC, FACC,  ORCID logo 
  • MARIANO GIORGI, 2  4  8  MTSAC  ORCID logo 
  • 1  Hypertension and Metabolic Unit, Hospital Universitario, Fundación Favaloro, Autonomous City of Buenos Aires, Argentina
  • 2  Cardiovascular Epidemiology and Prevention Council “Dr. Mario Ciruzzi" , de la Argentine Society of Cardiology .
  • 3  General Directorate of Education - Argentine Army .
  • 4  Argentine Lipid Society .
  • 5  Hospital Italiano de Buenos Aires, Autonomous City of Buenos Aires , Argentina
  • 6  Sanatorio Otamendi, Autonomous City of Buenos Aires , Argentina
  • 7  Hospital General de Agudos Dr. Juan A. Fernández, Autonomous City of Buenos Aires , Argentina
  • 8  Centro de Educación Médica e Investigaciones Clínicas “Norberto Quirno” (CEMIC), Autonomous City of Buenos Aires , Argentina
  • 9  Instituto Cardiovascular Lezica, Autonomous City of Buenos Aires , Argentina

Rev Argent Cardiol 2025;93:198-207. https://doi.org/10.7775/rac.v93.i3.20899

Received:  03/28/2025   Accepted:  05/09/2025   

Correspondence: Lorenzo M. Lobo. Correo electrónico: lorenzomartinlobo@hotmail.com

   


2025©Revista Argentina de Cardiología

Licencia Creative Commons  Creative Commons Atribution-NonCommercial-Sharelike 4.0 Internacional
This is an open-access article distributed under the terms of the Creative Commons Attribution License
 
 
ABSTRACT
Background:

Risk stratification in type 2 diabetes mellitus (DM2) patients constitutes a real clinical challenge.

Objectives:

1) To stratify cardiovascular risk in a population with DM2 in primary prevention, using different risk scores; 2) To describe the prevalence of atherosclerotic carotid plaque (ACP); 3) To analyze the correlation and concordance between the different risk scores; and 4) To evaluate the discriminative capacity of the different scores to identify the presence or absence of ACP.

Methods:

Observational, cross-sectional, multicenter study. Patients with DM2 in primary prevention without lipid-lowering treatment were included. The following risk scores were calculated: Framingham scores for cardiovascular disease (FrCVD) and for coronary heart disease (FrCHD), the 2018 ACC/AHA (AHA2018), PREVENT (AHA2023) and Diabetes-2 (S2-DBT) scores, and UKPDS scores for fatal coronary heart disease (UKPDS-FCHD), non-fatal coronary heart disease (UKPDS-NFCHD), fatal stroke (UKPDS-FS) and non-fatal stroke (UKPDS-NFS). Correlation and concordance between them were evaluated. The discriminative capacity of the different scores for PAC prediction was analyzed using ROC curves.

Results:

A total of 170 patients were included in the study, (mean age of 61.2±10.8 years, 58.8% men). The proportion of patients classified as at least high risk was 71.2%, 39.9%, 43.6%, 23.9%, 20.9%, 53.4%, 9.8% and 49.7% when the FrCVD, FrCHD, UKPDS- NFCHD, UKPDS-FCHD, UKPDS-NFS, AHA2018, AHA2023S and 2-DBT scores, respectively, were applied. No patient was classified with risk ≥20 % when using UKPDS-FS score. A good correlation was observed between all scores assessed (Spearman´s rho values between 0.64 and 0.98). The overall agreement between the scores was fair (Fleiss Kappa coefficient of 0.38). The prevalence of ACP was 50.6%, higher in the upper risk strata of all the scores evaluated. The analysis of the ROC curves showed a very good predictive capacity of all the scores for ACP detection.

Conclusion:

Cardiovascular risk was considerable for most of the scores evaluated, with an adequate correlation, but a fair concordance between the different scores. A very good predictive ability of all scores for ACP detection was observed.

Key words: Diabetes mellitus - Cardiovascular risk - Cardiovascular risk factors - Atherosclerosis - Carotid artery disease - Data correlation
 
 
RESUMEN
Introducción:

La estratificación de riesgo en el paciente con diabetes mellitus tipo 2 (DM2) constituye un verdadero desafío clínico.

Objetivos:

1) Estratificar el riesgo cardiovascular en una población con DM2 en prevención primaria, mediante la utilización de diversos puntajes de riesgo; 2) Describir la prevalencia de placa aterosclerótica carotídea (PAC) en las diferentes categorías de riesgo definidas por los puntajes utilizados; 3) Analizar la correlación y concordancia entre dichos puntajes; 4) Evaluar la capacidad discriminativa de las distintas puntuaciones para identificar la presencia o ausencia de PAC.

Material y métodos:

Estudio observacional, de corte transversal y multicéntrico. Se incluyeron pacientes con DM2 en preven- ción primaria sin tratamiento hipolipemiante. Se calcularon los siguientes puntajes de riesgo: Framingham para enfermedad cardiovascular (PF-CV) y para enfermedad coronaria (PF-Cor); ACC/AHA 2018 (P-AHA2018); PREVENT (P-AHA2023); SCORE2-Diabetes (PE2-DBT); y los puntajes UKPDS para enfermedad coronaria fatal (P-UKPDS-ECF), enfermedad coronaria no fatal (P-UKPDS-ECNF), accidente cerebrovascular fatal (P-UKPDS-ACVF) y no fatal (P-UKPDS-ACVNF). Se evaluó la correlación y concordancia entre ellos. Se analizó mediante curvas ROC la capacidad discriminativa de los distintos puntajes para la predicción de PAC

Resultados:

Se incluyeron 170 pacientes con edad media (desviación estándar) de 61,2 (10,8) años, 58,8 % hombres. La pro- porción de sujetos clasificados como de al menos alto riesgo fue de 71,2 %, 39,9 %, 43,6 %, 23,9 %, 20,9 %, 53,4 %, 9,8 % y 49,7 % cuando aplicamos los PF-CV, PF-Cor, P-UKPDS-ECNF, P-UKPDS-ECF, P-UKPDS-ACVNF, P-AHA2018, P-AHA2023

y PE2-DBT, respectivamente. Ningún paciente fue clasificado con un riesgo ≥20 % al utilizar el P-UKPDS-ACVF. Se observó una buena correlación entre todos los puntajes evaluados (valores de coeficiente rho de Spearman entre 0,64 y 0,98). La concordancia global entre los puntajes fue regular (coeficiente Kappa de Fleiss de 0,38). La prevalencia de PAC fue del 50,6 %, mayor en los estratos de mayor riesgo de todos los puntajes evaluados. El análisis de las curvas ROC mostró una muy buena capacidad predictiva de todos los puntajes para la detección de PAC.

Conclusión:

El riesgo cardiovascular observado fue considerable para la mayoría de las puntuaciones evaluadas, con una ade- cuada correlación, pero regular concordancia entre los distintos puntajes. Se observó una muy buena capacidad discriminativa de todos los puntajes para la detección de PAC.

Palabras clave: Diabetes mellitus - Riesgo cardiovascular - Factores de riesgo cardiovascular - Aterosclerosis - Enfermedad de las arterias carótidas - Correlación de datos
 
 

INTRODUCTION

In 2021, the global prevalence of type 2 diabetes mellitus (DM2) in people aged 20-79 years was estimated to be 10.5%, representing approximately 537 million individuals. (1) By 2045, the prevalence is projected to increase by 12.2%, equivalent to 783 million people. Healthcare costs associated with DM2, were estimated at $966 billion in 2021, and are projected to increase to more than $1 trillion in 2045. According to the Fourth National Survey of Risk Factors (2018), the prevalence of DM2 in Argentina, based on self- reporting, is 12.7%, which represents a significant increase compared with previous editions of that report. (2) This increase seems to be closely related to the sustained growth in the prevalence of obesity in the population.

Atherosclerotic cardiovascular disease continues to be the main cause of morbidity and mortality associated with DM2, (3) approximately doubling the risk of developing a wide variety of vascular diseases, independently of conventional risk factors (4). In this context, the CAPTURE study revealed that about one in three patients with DM2 has established cardiovascular disease. (5)

Based on early studies, the cardiovascular risk associated with DM2 was considered for many years to be equivalent to that of coronary heart disease. (6,7) However, with the emergence of new data, some authors have questioned this premise. (8,9) Although it is clear that patients with DM2 constitute a higher-risk population, it is likely that not all of them present the same level of vulnerability. In this context, the use of varous risk scores and additional tools, such as the detection of subclinical atheromatosis, could help to optimize cardiovascular risk stratification in this population. (10) A more accurate estimation of cardiovascular risk would allow better individualization of prevention strategies, focusing interventions of greater intensity (and greater costs) on subjects at higher risk. Although there are several tools used for cardiovascular risk estimation, and many of them have been evaluated in our population, (11) two new tools derived from the most widely used cohorts were published in mid-2023. The first is the updated 2023 American College of Cardiology (ACC) and American Heart Association (AHA) cardiovascular risk scores designed for the general population (called PREVENT, which we will refer to as AHA2023). The second is a specific tool for the diabetic population, developed in Europe, called SCORE2-Diabetes (which we will refer to as S2-DBT). (12-14) So far, we do not know how these tools would classify our patients with DM2, their relationship with other tools, and their ability to predict subclinical atherosclerotic disease.

Considering the above, the objectives of this work were: 1) To stratify cardiovascular risk in a population with DM2 in primary prevention, using different risk scores; 2) To describe the prevalence of atherosclerotic carotid plaque (ACP) in the different risk categories defined by the scores used; 3) To analyze the correlation and concordance between the different risk scores; 4) To evaluate the discriminative capacity of the different scores to identify the presence or absence of ACP.

METHODS

An observational, cross-sectional, multicenter cohort study was carried out, consecutively evaluating patients attending the cardiovascular prevention clinics of five centers located in the Autonomous City of Buenos Aires and the Province of Buenos Aires.

Patients with a diagnosis of DM2, defined by at least two fasting blood glucose values ≥126 mg/dl, blood glucose ≥200 mg/dl at 2 hours after the oral glucose tolerance test, or a glycosylated hemoglobin (HbA1c) value >6.4%, were included. (15,16) Patients with a history of acute myocardial infarction, coronary angioplasty, myocardial revascularization surgery, stroke, peripheral vascular disease, as well as patients receiving statin treatment were excluded. This last exclusion was intended to allow baseline risk stratification before any modification of the lipid profile, since this profile is a key component in the calculation of cardiovascular risk. Patients who were not receiving statins could have no medical indication or, more likely, could not be complying with the treatment despite having an indication for their use.

Laboratory data, clinical history, and usual medication were collected. Based on this information, the following 10-year cardiovascular risk scores were calculated: the Framingham score for cardiovascular disease (FrCVD) (17) and for coronary heart disease (FrCHD); (18) the ACC/AHA 2018 (AHA2018) (19) and AHA2023 risk scores; (12,13,20)

the S2-DBT for low-risk population score; (14) and the United Kingdom Prospective Diabetes Study (UKPDS) scores for fatal coronary heart disease (UKPDS-FCHD), nonfatal coronary heart disease (UKPDS-NFCHD), fatal stroke (UKPDS- FS), and nonfatal stroke (UKPDS-NFS). (21,22)

Given the availability of the method in all participating centers and the evidence supporting the presence of ACP as an independent predictor of coronary events, this tool was chosen as a surrogate marker of clinical atherosclerotic disease. (23) Ultrasound was used as a noninvasive method for ACP detection. Plaque characterization was based on the presence of the following criteria: 1) abnormal wall thickening (intima-media thickness >1.5 mm), 2) structural alteration (protrusion into the lumen or loss of alignment with the adjacent wall), and 3) abnormal wall echogenicity. The prevalence of ACP was compared between the different risk strata (quartiles) according to each of the scores used.

A Receiver Operating Characteristic (ROC) curve analysis was performed to determine the area under the curve (AUC), with the aim of evaluating the discriminative capacity of the different scores to identify the presence or absence of ACP. To establish the optimal cut-off point (OCP) for each score, the Youden index, defined as the maximum vertical distance between the ROC curve and the statistical probability line, was used.

Statistical analysis

Continuous variables with normal distribution were expressed as mean and standard deviation (SD,) and as median and interquartile range (IQR) for non-normal distribution. Comparison between groups was performed using Student's test in the case of normal distribution or the Mann-Whitney-Wilcoxon test when this condition was not met. Categorical variables were presented as percentages, and their comparison between groups was performed using the chi-square test. Due to the non-normal distribution of the scores evaluated, the correlation between the different risk scores was evaluated using Spearman's rho correlation coefficient, and was defined as null, poor, low, moderate, good or very good, if this coefficient was 0.0-0.20, 0.21-0.40, 0.41-0.60, 0.61-0.80, 0.81-1. The overall and pairwise concordance between risk scores was estimated according to the classification of patients into categories of "low/moderate" or "high/very high" cardiovascular risk, using the specific cut-off points defined for each score. The Fleiss Kappa coefficient was used to determine overall concordance and Cohen's Kappa coefficient for pairwise concordance, and were considered low, fair, moderate, good and very good, if the coefficients were <0.20, 0.21-0.40, 0.41-0.60, 0.61-0.80, or 0.81-1, respectively. A value of two tailed p<0.05 was considered statistically significant. Statistical analysis was performed using R v 4.4.2 software. (24)

Ethical considerations

The present study was performed in accordance with the ethical standards of the declaration of Helsinki (25), and the law on personal data protection 25326. (26)

RESULTS

A total of 170 patients were included in the study, mean age of 61.2 (10.8) years, 58.8% men. Mean HbA1c was 7.1 (1.3) and 14.9% were receiving insulin treatment. Median time of DM2 evolution was 5 (2-10) years. A total of 31.7% patients were overweight, 51.1% had some degree of obesity, and only 1 out of 5 was at a healthy weight. A total of 73.4 % had low-density lipoprotein cholesterol (LDL-C) levels above 100 mg/dl. In addition, 64.1 % of the patients had hypertension, 18.8 % were smokers and 3.6 % had a history of atrial fibrillation. Table 1 shows population characteristics. Median (IQR) values of FrCHD, FrCVD, S2-DBT, AHA2018, and AHA2023 scores were 18 (10.3-27.0), 29 (16.5-47.8), 9.8 (6.7-13.0), 21.7 (9.5-37.0), and 10.0 (5.8-15.0), respectively. Similarly, according to the UKPDS scores, the median values for coronary heart disease (fatal and nonfatal), fatal coronary heart disease, stroke (fatal and nonfatal), and fatal stroke were 18 (10.5-29.1), 10.7 (5.2-19.3), 8.0 (3.7-15.9), and 1.1 (0.5-2.2), respectively.

Table 1

Total population characteristics, stratified by sex.

Continuous variables, mean (SD) or median (IQR)Overall Male Female p
n=170n = 100n = 70
Age, years61.2 (10.8)61.0 (11.3)61.5 (10.1)0.761
Total cholesterol, mg/dL200.7 (36.2)202.7 (40.2)197.8 (29.6)0.384
LDL-C, mg/dL121.1 (34.1)126.3 (37.5)113.9 (27.4)0.021
HDL-C, mg/dL45.7 (13.4)41.8 (10.4)51.3 (15.2)<0.001
Triglycerides, mg/dL172.4 (101.2)184.3 (115.7)155.4 (73.3)0.067
Body mass index, Kg/m230.6 (5.7)29.9 (5.2)31.7 (6.1)0.043
HbA1c, %7.1 (1.3)7.1 (1.2)6.9 (1.4)0.328
DM2 evolution, years5 (2-10)6 (2-10)4 (2-8)0.101
Categorical variables, n (%)n = 170n = 100n = 70
Hypertension109 (64.1)70 (70.0)39 (55.7)0.084
Active smoking32 (18.8)25 (25.0)7 (10.0)0.001
Atrial fibrillation6 (3.6)6 (6.1)0 (0.0)0.097
Family history of early coronary artery disease30 (17.6)23 (23.0)7 (10.0)0.047
ACP86 (50.6)60 (60.0)26 (37.1)0.005
Insulinized25 (14.9)18 (18.2)7 (10.1)0.223

ACP: atherosclerotic carotid plaque; DM2: type 2 diabetes mellitus; HbA1c: glycosylated hemoglobin; HDL-C: high-density lipoprotein cholesterol; IQR: interquartile range; LDL-C: low-density lipoprotein cholesterol; SD: standard deviation

The proportion of subjects classified as at least high risk was 71.2%, 39.9%, 53.4%, 9.8%, and 49.7% when applying the FrCVD, FrCHD, AHA2018, AHA2023, and S2-DBT scores, respectively. Regarding the UK-PDS models, the proportion of patients so categorized was 43.6%, 23.9% and 20.9% when using UKPDS- NFCHD, UKPDS-FCHD and UKPDS-NFS scores, respectively. Of note, no patient was classified with a risk ≥ 20% when applying the UKPDS-FS score.

A significant correlation was observed between all the scores assessed, with Spearman´s rho coefficient values ranging from 0.64 to 0.98 (Figure 1A). However, despite this correlation, the overall concordance between scores was fair, with a Fleiss Kappa coefficient of 0.38. Pairwise agreement was evaluated for all the scores, with the exception of the UKPDS-FS score, because it classified all patients as "low/moderate risk" and the categories could not be compared. Cohen's Kappa coefficients are presented in Figure 1B.

Fig 1

Correlation and concordance between the different scores assessed. A: Correlation coefficient between the different cardiovascular risk scores. B: Overall and pairwise concordance of the different cardiovascular risk scores. The UKPDS-FS score was not evaluated since no patient was categorized as high risk. For score abbreviations see the Methods section

Lobo-en-gf1.jpg 

Overall, the prevalence of ACP was 50.6%, with a higher proportion in men than in women (60% vs. 37.1%, p=0.005). This prevalence increased progressively in the higher risk strata (quartiles) in all scores assessed. Although the highest prevalence was recorded in the upper quartile with the UKPDS-NFCHD and AHA2023 scores (95.2% and 95.1%, respectively), the range of variation within this quartile was relatively narrow (81.4% - 95.2%) (Figure 2). Patients with the presence of ACP were significantly older 62.7(7.7) vs. 55.1 (10.0) years, p <0.001, and with a lower proportion of women (30.2% vs. 52.4%, p=0.005). In addition, they had more frequent hypertension, smoking habit and a longer time of DM2 evolution compared with the group without ACP. All cardiovascular risk scores assessed showed significantly higher mean values in patients with ACP (Table 2).

Fig. 2

Prevalence of plaque, according to the cardiovascular risk quartiles of the different scores evaluated. For score abbreviations see the Methods section.

Lobo-en-gf2.jpg 

Table 2

Patient characteristics according to the presence or absence of ACP

Continuous variables, mean (SD) or median (IQR).Without ACPWith ACPp
n=84n=86
Age, years55.1 (10.0)67.2 (7.7)<0.001
Total cholesterol, mg/dl199.0 (35.9)202.4 (36.6)0.542
LDL-C, mg/dl115.7 (32.7)126.4 (34.9)0.043
HDL-C, mg/dl47.3 (14.7)44.2 (11.8)0.13
Triglycerides, mg/dl185.6 (125.2)159.4 (68.6)0.092
Body mass index, Kg/m231.3 (5.8)30.0 (5.5)0.158
HbA1c, %6.9 (1.4)7.2 (1.1)0.138
DM2 evolution, years4 (1.25-6)8 (3-12)<0.001
FrCVD,%19.6 (9.4-30.4)42.4 (27.4-56,8)<0.001
FrCHD,%13,0 (8.0-20.0)24.5 (16.0-31.0)<0.001
AHA2018,%10.1 (4.6-21.2)34.4 (21.6-48.7)<0.001
UKPDS-NFCHD,%10.7 (6.4-16.3)26.5 (18.6-44.1)<0.001
UKPDS-NFS,%4.3 (1.6-6.8)14.5 (8.3-29.4)<0.001
UKPDS-FCHD,%5.5 (2.5-10.4)17.8 (10.9-34.9)<0.001
UKPDS-FS,%0.6 (0.2-1.0)2.0 (1.20-4.0)<0.001
S2-DBT,%6.8 (4.4-9.5)12.9 (10.1-16.5)<0.001
AHA2023,%6.0 (3.8-9.4)13.9 (10.3-19.1)<0.001
Categorical variables, n (%)Without ACP n=84With ACP n=86p
Women44 (52.4)26 (30.2)0.005
Hypertension41 (48.8)68 (79.1)<0.001
Active smoking10 (11.9)22 (25.6)0.002
Atrial fibrillation1 (1.2)5 (5.8)0.235
Family history of early coronary artery disease15 (17.9)15 ( 17.4)1
Insulinized10 (12.2)15 (17.4)0.46

ACP: atherosclerotic carotid plaque; DM2: type 2 diabetes mellitus; HDL-C: high-density lipoprotein cholesterol; IQR: interquartile range. LDL-C: low-density lipoprotein cholesterol; SD: standard deviation. For score abbreviations see the Methods section.

Analysis of ROC curves showed a very good predictive ability of all scores for plaque detection. The UK-PDS scores presented practically identical AUC values (≈ 0.88). Among the remaining scores, those that best classified patients were the AHA2018, AHA2023, and S2-DBT, with AUC values of 0.86, 0.86, and 0.87, respectively. As shown in Figure 3, the OCPs varied according to the score evaluated.

Fig. 3.

ROC curves, AUC and OCP of the different cardiovascular risk scores evaluated. For score abbreviations see the Methods section.

Lobo-en-gf3.jpg 

DISCUSSION

The identification of patients at risk of developing cardiovascular events represents one of the most complex challenges in clinical practice. Over time, various tools have been proposed to estimate the risk of future events in patients with DM2, although new cardiovascular risk calculators have recently been introduced. In our study, we compared eight predictive scales in a group of patients with DM2 without history of cardiovascular disease and analyzed their relationship with the presence of ACP.

A key point for discussion is the role of risk scores in patients with DM2 in primary prevention. The 2018 AHA/ACC guidelines for lipid management recommend administration of moderate-intensity statins in patients aged 40-75 years with DM2, and no cardiovascular history. (19) However, they suggest stratifying cardiovascular risk (AHA18S) to administer high-intensity statins in those at higher risk. An expert consensus on the role of non-statin therapies, published in 2022 by the same scientific societies, recommends in this population the administration of high-intensity statins when the calculated cardiovascular risk is equal to or greater than 7.5% at 10 years, and the addition of ezetimibe when the risk exceeds 20% and LDL-C is >70 mg/dL. (27) On the other hand, the 2023 European guidelines for the management of cardiovascular disease in patients with DM2 suggest using the S2DBT to classify patients into the following categories: low risk (<5%), moderate risk (5-10%), high risk (>10% and <20%), and very high risk (≥20%). (28) Lipid-lowering medication, the choice of antidiabetic drugs, and therapeutic targets vary according to the risk of each patient. Similarly, the latest Cardiovascular Prevention Consensus of the Argentine Society of Cardiology (SAC) follows this approach, endorsing cardiovascular risk stratification as the basis for defining the type and intensity of treatments in the patient with DM2. (29) Our study revealed that a significant proportion of persons with DM2 were classified as at least high cardiovascular risk according to most of the scores evaluated. However, this proportion varies, being lower when scores assessing a reduced number of clinical events are analyzed and, even more, if only fatal events are considered. Such is the case of the UKPDS-FS, where no patient showed a risk ≥20%. On the other hand, the scores used do not always consider the same variables. A particular case is that of the-AHA2018 and AHA2023 scores, where the difference in the classification of patients as high risk was notably greater in the former (50.4% vs. 9.8%). Unlike the 2018 calculator, PREVENT includes body mass index, estimated glomerular filtration rate, and optionally HbA1c and urinary albumin-to-creatinine ratio.

In addition, PREVENT omits ethnicity, which was included in the 2018 calculator. In fact, in other populations it has also been observed that the AHA18S tends to overestimate risk compared to the AHA2023. (30)

The concepts of concordance and correlation are widely used to assess the association between variables, but, although related, they represent different approaches. (31) Correlation focuses on the relationship between observed changes in two variables, i.e., how one varies as a function of the other. In contrast, concordance focuses on the degree of coincidence between two methods of measuring or classifying the same variable. Thus, while correlation measures association, concordance evaluates the degree of agreement between the results. In this sense, our study showed a high correlation between the different models analyzed. Previously, our research group had reported similar results in a population with DM2, though without including the new cardiovascular risk scales. (32) However, in our study, the concordance in the classification of patients as "high risk" was low or moderate among the different scores evaluated. In the general population, several studies have shown that the application of different cardiovascular risk prediction models in the same group of patients can generate discrepancies in risk classification in about 30% of cases. (33) In particular, when comparing the Framingham model with the European SCORE model, the concordance reported was mild to moderate. (34,35) Similar data have been reported in our region. For example, in Colombia, the kappa coefficient of concordance between the Framingham model and SCORE for the classification of patients as high risk was 0.28, indicating low concordance. (36) Similarly, in Peru, limited agreement was observed between the AHA/ACC equations and six different cardiovascular risk estimation scales. (37)

A recent analysis by Dziopa et al. compared the performance of 22 cardiovascular risk models, including both those designed specifically for persons with DM2 and others originally developed for the general population. The study evidenced remarkable variability in the predictive ability of these models. Surprisingly, the results indicated that models based on the general population did not necessarily underperform compared with DM2-specific models. (38) Even the non-specific scores included a small percentage of patients with DM2 in their derivation cohorts (e.g., FrCVD score 5%, FrCHD score 5.2%, and AHA2018 score between 1% and 14%, depending on the cohort evaluated). This unexpected result highlights that DM2-specific models do not necessarily outperform general models in predicting cardiovascular events. In other words, these findings challenge the assumption that incorporating DM2-related variables-such as HbA1c and disease duration-always improves risk predictive ability. (39) After adjusting for differences in population characteristics and baseline risk, recalibration significantly improved the performance of many models. This emphasizes the importance of making local adaptations of widely used models (recalibration). (40) Overall, the global prevalence of ACP in people aged 30-79 years in 2020 was estimated to be 21.1%, equivalent to approximately 816 million people affected, with 59% increase since 2000. (41) As expected, DM2 proved to be one of the risk factors most closely associated with the increased likelihood of developing ACP. In our study, we found that approximately 1 in 2 patients with DM2 had ACP. These results are consistent with previous reports. For example, a study conducted in Taiwan (42) reported a prevalence of 53.6%, whereas another conducted in China found 44.9%. (43) In the latter case, and as in our study, the authors reported that subjects with ACP were older, had longer evolution of DM2, and had a higher prevalence of hypertension. Finally, and in line with the current findings, the study previously conducted by this group of investigators showed a prevalence of ACP of 51%. (32) These findings have notable clinical implications. According to the SAC Consensus on the management of patients with DM2 and cardiovascular disease in patients with low or intermediate risk (<20%), could be useful to screen for subclinical atheromatosis, since its presence would allow the patient with DM2 to be recategorized as very high risk. (44) Furthermore, this finding justifies the intensification of lipid-lowering treatment, also prioritizing the use of antidiabetic drugs with proven cardiovascular efficacy.

Finally, in our investigation, all the scores evaluated demonstrated a very good ability to predict the presence of ACP, with AUC above 0.75. However, the OCP varied considerably among the different scores evaluated. Some showed values close to the high-risk threshold (20%) established by the guidelines recommending their use, such as the FCVS and FCDS. Similarly, S2-DBT score presented an OCP close to 10%, the threshold considered by the respective guidelines to define high cardiovascular risk. In contrast, other scores, such as the AHA2018 score and, especially, the AHA2023 score showed significantly lower OCP than the commonly used high-risk thresholds, suggesting that many patients with scores that were not as high could nevertheless present ACP.

This study has several limitations. First, it was a cross-sectional study with a small number of patients. Consequently, our findings should be confirmed in larger studies. Secondly, all participants were recruited from outpatient cardiovascular prevention clinics belonging to cardiology centers, which could have introduced a selection bias. Furthermore, this conditions its extrapolation to daily health care practice. Finally, this study did not aim to evaluate the validity of the risk scores analyzed in our country. However, we consider that our findings highlight the difficulties inherent to cardiovascular risk stratification in this population, as well as the relevance of having local adaptations of the predictive models usually employed.

CONCLUSIONS

The observed cardiovascular risk was considerable according to most of the scores assessed; however, risk stratification was heterogeneous, with adequate correlation but only modest concordance between the different scores. In other words, these findings reflect that our patients with DM2 are not uniformly classified when applying different predictive tools. The prevalence of ACP was significantly increased in strata with higher estimated risk. Understanding the relationship between the presence of ACP and different estimation tools could improve the accuracy of cardiovascular risk assessment in patients with DM2 and optimize treatment in those most likely to have an event.

 

Funding

None.

Conflicts of interest

None declared. (See authors' conflict of interests forms on the web).

 
 

REFERENCES

1. Sun H, Saeedi P, Karuranga S, Pinkepank M, Ogurtsova K, Duncan BB, et al. IDF Diabetes Atlas: global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045. Diabetes Research and Clinical Practice 2022;183:109119. https://doi.org/1016/j.diabres.2021.109119.
2. National Directorate of Health Promotion and Control of Chronic Noncommunicable Diseases. Fourth National Survey of Risk Factors, 2018 [Internet]. [cited April 26, 2025]. Accessed April 26, 2025. https://www.indec.gob.ar/indec/web/Nivel4-Tema-4-32-68.
3. Leon BM. Diabetes and cardiovascular disease: epidemiology, biological mechanisms, treatment recommendations and future research. WJD 2015;6:1246. https://doi.org/10.4239/wjd.v6.i13.1246
4. Sarwar N, Kondapally Seshasai S, Gobin R, Kaptoge S, Di Angelantonio E, Ingelsson, E, et al. Diabetes Mellitus, Fasting Glucose, and Risk of Cause-Specific Death. The Emerging Risk Factors Collaboration. N Engl J Med 2011;364:829-41. https://doi.org/10.1056/NEJMoa1008862.
5. Mosenzon O, Alguwaihes A, Leon JLA, Bayram F, Darmon P, Davis TME, et al. CAPTURE: a multinational, cross-sectional study of cardiovascular disease prevalence in adults with type 2 diabetes across 13 countries. Cardiovasc Diabetol 2021;20:154. https://doi.org/10.1186/s12933-021-01344-0
6. Haffner SM, Lehto S, Rönnemaa T, Pyörälä K, Laakso M. Mortality from Coronary Heart Disease in Subjects with Type 2 Diabetes and in Nondiabetic Subjects with and without Prior Myocardial Infarction. N Engl J Med 1998;339:229-34. https://doi.org/10.1056/NEJM199807233390404.
7. Juutilainen A, Lehto S, Rönnemaa T, Pyörälä K, Laakso M. Type 2 Diabetes as a "Coronary Heart Disease Equivalent". Diabetes Care. 2005;28:2901-7. https://doi.org/10.2337/diacare.28.12.2901
8. Evans JMM. Comparison of cardiovascular risk between patients with type 2 diabetes and those who had had a myocardial infarction: cross sectional and cohort studies. BMJ 2002;324:939. https://doi.org/10.1136/bmj.324.7343.939
9. Bulugahapitiya U, Siyambalapitiya S, Sithole J, Idris I. Is diabetes a coronary risk equivalent? Systematic review and meta-analysis. Diabetic Medicine 2009;26:142-8. https://doi.org/10.1111/j.1464-5491.2008.02640.x
10. Damaskos C, Garmpis N, Kollia P, Mitsiopoulos G, Barlampa D, Drosos A, et al. Assessing Cardiovascular Risk in Patients with Dia- betes: An Update. CCR 2021;16:266-74. https://doi.org/10.2174/1573403X15666191111123622
11. Rubinstein AL, Irazola VE, Poggio R, Bazzano L, Calandrelli M, Lanas Zanetti FT, et al. Detection and follow-up of cardiovascular disease and risk factors in the Southern Cone of Latin America: the CESCAS I study. BMJ Open 2011;1:e000126. https://doi.org/10.1136/bmjopen-2011-000126. https://doi.org/10.1136/bmjo-pen-2011-000126
12. Larkin H. What to Know About PREVENT, the AHA's New Cardiovascular Disease Risk Calculator. JAMA 2024;331:277. https://doi.org/10.1001/jama.2023.25115
13. Khan SS, Matsushita K, Sang Y, Ballew SH, Grams ME, Surapaneni A, et al. Development and Validation of the American Heart Association's PREVENT Equations. Circulation 2024;149:430-49. https://doi.org/10.1161/CIRCULATIONAHA.123.067626
14. Pennells L, Østergaard HB, Read SH, Carinci F, Franch-Nadal J, Petitjean C, et al. SCORE2-Diabetes: 10-year cardiovascular risk estimation in type 2 diabetes in Europe. SCORE2-Diabetes Working Group and the ESC Cardiovascular Risk Collaboration. Eur Heart J 2023;44:2544-56. https://doi.org/10.1093/eurheartj/ehad260.
15. Maseli M del C, Llanos I, Lucarelli C, Fenili C, Ruibal G, Valdez. Opinions and recommendations of the Argentine Diabetes Society Hemoglobin A1c. Rev Soc Argent Diabetes. 2023;57(1):20-3.
16. National Ministry of Health. Guía de Práctica Clínica Nacional sobre Prevención, Diagnóstico y Tratamiento de la Diabetes Mellitus tipo 2, 2019. Buenos Aires, Argentina. https://argentina.org.ar.
17. D'Agostino RB, Vasan RS, Pencina MJ, Wolf PA, Cobain M, Massaro JM, et al. General Cardiovascular Risk Profile for Use in Prima- ry Care: The Framingham Heart Study. Circulation 2008;117:743-53. https://doi.org/10.1161/CIRCULATIONAHA.107.699579
18. Wilson PWF, D'Agostino RB, Levy D, Belanger AM, Silbershatz H, Kannel WB. Prediction of Coronary Heart Disease Using Risk Factor Categories. Circulation 1998;97:1837-47. https://doi.org/10.1161/01.CIR.97.18.1837
19. 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. J Am Coll Cardiol 2019;73:3168-209. https://doi.org/10.1016/j.jacc.2018.11.002.
20. Khan SS, Coresh J, Pencina MJ, Ndumele CE, Rangaswami J, Chow SL, et al. Novel Prediction Equations for Absolute Risk Assessment of Total Cardiovascular Disease Incorporating Cardiovascular-Kidney-Metabolic Health: A Scientific Statement From the American Heart Association. Circulation 2023;148:1982-2004. https://doi.org/10.1161/CIR.0000000000001191
21. Stevens RJ, Kothari V, Adler AI, Stratton IM, United Kingdom Prospective Diabetes Study (UKPDS) Group. The UKPDS risk engine: a model for the risk of coronary heart disease in Type II diabetes (UKPDS 56). Clin Sci (Lond). 2001;101:671-9.
22. Kothari V, Stevens RJ, Adler AI, Stratton IM, Manley SE, Neil HA, et al. UKPDS 60: risk of stroke in type 2 diabetes estimated by the UK Prospective Diabetes Study risk engine. Stroke 2002;33:1776-81. https://doi.org/10.1161/01.str.0000020091.07144.c7
23. Nambi V, Chambless L, Folsom AR, He M, Hu Y, Mosley T, et al. Carotid Intima-Media Thickness and Presence or Absence of Plaque Improves Prediction of Coronary Heart Disease Risk. J Am Coll Cardiol 2010;55:1600-7. https://doi.org/10.1016/j.jacc.2009.11.075 .
24. R Core Team. R: A Language and Environment for Statistical Computing [Internet]. Vienna, Austria: R Foundation for Statistical Computing; 2024. https://www.R-project.org/.
25. World Medical Association. World Medical Association Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects. JAMA. 2013;310 :2191-2194. https://doi.org/10.1001/jama.2013.281053
26. https://servicios.infoleg.gob.ar/infolegInternet/anexos/60000-64999/64790/texact.htm
27. Lloyd-Jones DM, Morris PB, Ballantyne CM, Birtcher KK, Covington AM, DePalma SM, et al. 2022 ACC Expert Consensus Decision Pathway on the Role of Nonstatin Therapies for LDL-Cholesterol Lowering in the Management of Atherosclerotic Cardiovascular Disease Risk. J Am Coll Cardiol 2022;80:1366-418. https://doi.org/10.1016/j.jacc.2022.07.006
28. Marx N, Federici M, Schütt K, Müller-Wieland D, Ajjan RA, Antunes MJ, et al. 2023 ESC Guidelines for the management of cardiovascular disease in patients with diabetes. Eur Heart J 2023;44:4043-140. https://doi.org/10.1093/eurheartj/ehad192.
29. Giunta G, Lavalle Cobo A, Brandani L, Lobo M, Forte E, Masson G, et al. Consensus on Cardiovascular Prevention - Summary version. Rev Argent Cardiol 2023;91(S3):1-15. http://dx.doi.org/10.7775/rac.es.v91.s3
30. Anderson TS, Wilson LM, Sussman JB. Atherosclerotic Cardiovascular Disease Risk Estimates Using the Predicting Risk of Cardiovascular Disease Events Equations. JAMA Intern Med 2024;184:963. https://doi.org/10.1001/jamainternmed.2024.1302
31. Liu J, Tang W, Chen G, Lu Y, Feng C, Tu XM. Correlation and agreement: overview and clarification of competing concepts and measures. Shanghai Arch Psychiatry 2016;28:115-20. https://doi.org/10.11919/j.issn.1002-0829.216045
32. Masson W, Francesca SD, Molinero M, Siniawski D, Mulassi A, Morales FE, et al. Estimation of cardiovascular risk and detection of subclinical carotid atheromatosis in patients with diabetes without a history of cardiovascular disease. Arch Endocrinol Metab 2017;61:122-9. https://doi.org/10.1590/2359-3997000000234.
33. Allan GM, Nouri F, Korownyk C, Kolber MR, Vandermeer B, Mc-Cormack J. Agreement Among Cardiovascular Disease Risk Calculators. Circulation 2013;127:1948-56. https://doi.org/10.1161/CIRCULATIONAHA.112.000412
34. Fornasini M, Brotons C, Sellarès J, Martinez M, Galán ML, Sáenz I, et al. Consequences of using different methods to assess cardiovas- cular risk in primary care. Family Practice 2006;23:28-33. https://doi.org/10.1093/fampra/cmi092
35. González C, Rodilla E, Costa JA, Justicia J, Pascual JM. Comparison between the Framingham algorithm and the SCORE algorithm in the calculation of cardiovascular risk in subjects aged 40-65 years. Medicina Clinica 2006;126:527-31. https://doi.org/10.1157/13087144.
36. Muñoz V OM, Ruiz Morales ÁJ, Mariño Correa A, Bustos C. MM. Concordance between the SCORE and Framingham models and the AHA/ACC equations as cardiovascular risk assessors. Rev Colomb Cardiol 2017;24:110-6. https://doi.org/10.1016/j.rccar.2016.06.013.
37. Bazo-Alvarez JC, Quispe R, Peralta F, Poterico JA, Valle GA, Burroughs M, et al. Agreement Between Cardiovascular Disease Risk Scores in Resource-Limited Settings: Evidence from 5 Peruvian Sites. Critical Pathways in Cardiology: A Journal of Evidence-Based Medicine 2015;14:74-80. https://doi.org/10.1097/HPC.0000000000000045.
38. Dziopa K, Asselbergs FW, Gratton J, Chaturvedi N, Schmidt AF. Cardiovascular risk prediction in type 2 diabetes: a comparison of 22 risk scores in primary care settings. Diabetologia 2022;65:644-56. https://doi.org/10.1007/s00125-021-05640-y
39. Zanfirescu RL, Anghel L, Tudurachi BS, Clement AM, Z voi A, Benchea LC, et al. Improved ASCVD Screening in Diabetes: a Focus on Scoring Models and Detection Techniques. Romanian Journal of Internal Medicine [Internet]. 2025 [cited 2025 May 5]; https://doi.org/10.2478/rjim-2025-0009.
40. Van Der Leeuw J, Van Dieren S, Beulens JWJ, Boeing H, Spijkerman AMW, Van Der Graaf Y, et al. The validation of cardiovascular risk scores for patients with type 2 diabetes mellitus. Heart 2015;101:222-9. https://doi.org/10.1136/heartjnl-2014-306068
41. Song P, Fang Z, Wang H, Cai Y, Rahimi K, Zhu Y, et al. Global and regional prevalence, burden, and risk factors for carotid atherosclerosis: a systematic review, meta-analysis, and modelling study. The Lancet Global Health 2020;8(5):e721-9. https://doi.org/10.1016/S2214-109X(20)30117-0.
42. Wu TW, Chou CL, Cheng CF, Lu SX, Wang LY. Prevalences of diabetes mellitus and carotid atherosclerosis and their relationships in middle-aged adults and elders: a community-based study. J Formosan Med Assoc. 2022;121:1133-40. https://doi.org/10.1016/j.jfma.2021.10.005.
43. Chen J, Li W, Cao J, Lu Y, Wang C, Lu J. Risk factors for carotid plaque formation in type 2 diabetes mellitus. J Transl Med 2024;22:18. https://doi.org/10.1186/s12967-023-04836-7.
44. Giorgi, MA, Litwak LE, Sanabria H, Lavalle-Cobo A, Musso C, Salmeri E, et al. Consensus on the Management of Patients with Diabetes Mellitus and Cardiovascular Pathology. Argentine Society of Cardiology (SAC). Rev Argent Cardiol 2020;88:1-58.

 
 

HTML generado desde un xml-jats a través de la plantilla jats-rac.xsl  por  RevisCiencia.  El motor utilizado es Apache Software Foundation (Xalan XSLTC).