Introduction

Cardiovascular risk (CVR) refers to the individual's probability of developing cardiovascular disease (CVD). It involves the occurrence of events such as myocardial infarction, cerebral hemorrhage, and embolism¹, usually caused by obstructions that restrict blood flow to the heart or brain. These obstructions, in turn, result from the formation of atherosclerotic plaque or fatty deposits in the walls of blood vessels².

When talking about CVR, it is important to recognize that arterial hypertension (AHT) and type 2 diabetes mellitus (DM2) are the most relevant risk factors for cardiovascular events. In Latin America, the prevalence of hypertension is reported to range between 30% and 50%, while DM2 ranges between 8% and 13%³. In the department of Putumayo in the year 2020, the crude incidence rates per 100,000 inhabitants were 1.79 for AHT (n=354), 0.99 (n=354) for DM2, and 2.86 (n=1019) for chronic kidney disease, taking into account that mortality rates per 100,000 inhabitants for these conditions were 93.14 (n=332), 33.10 (n=118), and 46.01 (n=164), respectively⁴.

Likewise, a frequently used tool recognized worldwide for measuring CVR is available: the Framingham scale, which estimates the 10-year risk in persons over 30 years. It classifies them as low risk if the sum of the scores is below 10%, moderate risk if it falls between 10% and 20%, and high risk if it is >20%⁵. In Colombia, this scale has been adapted by multiplying the obtained score by 0.75⁶. This tool becomes crucial in programs for patients with chronic conditions, as it helps identify CVR in a population at higher risk of developing cardiovascular disease (CVD) and requiring hospitalization due to their baseline condition, leading to an increased need for medical care⁷.

On the other hand, despite Colombia’s high-coverage health system, barriers to healthcare access persist, particularly for those living in rural areas of the country⁸,⁹. The literature reports that the health situation in rural areas is more complex and is influenced by factors such as nutritional imbalances, high fertility rates, excessive workloads, poor environmental conditions¹⁰, and difficult access to education¹¹. Consequently, these factors underscore the relevance of sociodemographic conditions in determining CVR, alongside the previously mentioned clinical factors.

Therefore, assessing potential differences in CVR between rural and urban areas is important, considering the different sociodemographic and clinical characteristics. This assessment allowed us to answer the following question: What factors are associated with high cardiovascular risk in patients from rural areas compared to those residing in urban areas who are enrolled in a chronic disease program at health institutions in Putumayo?

Materials and Methods

Study design and population: A cross-sectional study with an analytical component. The study population comprised adult patients enrolled in a chronic conditions program in one of the health service provider institutions (IPS, for its acronym in Spanish) in the department of Putumayo during the second semester of 2023.

Eligibility criteria: Patients aged 30 to 74 years were included. Pregnant women and individuals with incomplete Framingham equation variables in the database, preventing cardiovascular risk calculation, were excluded.

Sample size and sampling: The sample size was calculated for a logistic regression using the formula

where n represents the required sample size, k is the number of variables included in the model (6 in this case), which was multiplied by 10 cases per variable, and finally divided by the prevalence of cardiovascular disease (the dependent variable) reported as 5.88%¹². Based on these parameters, the estimated sample size was 1,190 patients.

The sample was selected through simple random sampling, from a database provided by two participating IPS from urban low- and medium-complexity care centers, with an eligible population of 6,377 in the two municipalities. No missing values were obtained in the sample data.

Statistical analysis: The analysis was conducted using Stata version 18 with the total sample (1,190). The geographic area of residence (rural or urban), educational level, ethnicity, type of health affiliation, cigarette smoking, history of AHT, history of DM, presence of dyslipidemia, and use of antihypertensive treatment were described using absolute and relative frequencies. According to the Framingham scale, age and CVR scores were reported as medians and interquartile ranges (IQR) after assessing normality with the Shapiro-Wilk test.

CVR analyses were performed considering two categories according to the Framingham scale score: low/moderate and high. First, a comparison between the proportion of these CVR categories and area of residence (urban versus rural) was made using a χ2 test. Subsequently, bivariate analyses were conducted using logistic regression, considering CVR levels and each independent variable. Variables with a p≤0.25 in the bivariate analysis, along with those deemed clinically relevant, were included in a multivariate logistic regression model. For variables that were not statistically significant (educational level and ethnicity), the power calculations were performed, yielding 100% for both; consequently, for this sample, these variables were not associated.

A multivariate logistic regression was conducted, confirming that all variables were significant for inclusion in the multivariate model. Subsequently, possible confounding variables were assessed using the coefficient change method, and interactions were examined by incorporating interaction terms into the logistic regression model. Finally, regression assumptions and model diagnostics were evaluated using the Hosmer-Lemeshow test, which yielded a p-value of 0.96, indicating a good model fit¹³. The complete dataset collected is freely accessible for consultation on Mendeley Data¹⁴.

The study protocol was submitted for approval to the Ethics Committee of the Fundación Universitaria de Ciencias de la Salud (FUCS) and classified as “no-risk research”¹⁵ according to Resolution 8430 of 1993 issued by the Colombian Ministry of Health. Additionally, to safeguard privacy and confidentiality, subjects identified during sample construction were assigned sequentially ordered codes in compliance with the guidelines established in Law 1581 of 2012 to protect personal data¹⁶. In addition, this study was approved by the Research Ethics Committee of the Universidad Cooperativa de Colombia as per Minute No. 6, dated 06/22/2023.

Results

Results from 1,190 patients identified during the study period are presented. The median age of the study population was 59.5 years (IQR 53-66), with a median of 59 years in urban areas (IQR 52-66) and 60 years in rural areas (IQR 53-65). Regarding sex, a lower proportion of women was observed in urban areas (67.09%, n=479) compared to rural areas (71.22%, n=339). In terms of educational level, most of the population completed primary education, with a slight numerical difference between urban (66.95%, n=478) and rural areas (72.27%, n=344). Most patients had subsidized health affiliation in both urban (97.76%, n=698) and rural areas (98.11%, n=467). In addition, a comparison of variables and area (rural or urban) revealed statistically significant differences in educational level and ethnicity. Other demographic characteristics are shown in Table 1.

Table 1

Sociodemographic variables of the study population

*IQR: Interquartile range; ** χ2 test for categorical variables and Mann-Whitney U test.

Regarding the study population's risk factors and clinical conditions, the proportion of tobacco dependence is low in urban (1.96%, n=14) and rural areas (1.47%, n=7). The proportion of dyslipidemia was similar in both areas (urban 35.15%, n=251 and rural 36.76%, n=175). Type II diabetes was proportionally more frequent in urban areas (64.99%, n=464) than in rural areas (67.65%, n=322). The AHT proportion was high in both urban (91.32%, n=652) and rural (89.50%, n=426) areas. In addition, a comparison of clinical variables by area (rural or urban) revealed statistically significant differences in diastolic blood pressure, antihypertensive treatment, glycosylated hemoglobin, total cholesterol, and LDL cholesterol. Other clinical characteristics are shown in Table 2.

Table 2

Clinical variables

*IQR: Interquartile range; HCTZ: Hydrochlorothiazide; ACE: Angiotensin-Converting Enzyme; ARA: Angiotensin Receptor Antagonists. HDL: High-Density Lipoprotein LDL: Low-Density Lipoprotein

The most frequent CVR level for the study population was high, with a numerically higher proportion in the urban areas (87.95%) compared to the rural areas (83.61%). When comparing CVR level with geographic areas of residence, a statistically significant difference was found (p=0.033) (See Figures 1).

Figure 1
Cardiovascular risk in patients living in rural areas vs. population living in urban areas
* χ2 test with a p-value = 0.033.

Finally, within the results observed in the multiple logistic regression, the final model included the following variables: primary education (OR: 0.68, CI 95%: 0.38–1.24), secondary education (OR: 0.88, CI 95%: 0.42–1.83), higher education (OR: 0.33, CI 95%: 0.13–0.82), rural area (OR: 0.66, CI 95%: 0.47–0.94), subsidized health affiliation (OR: 6.58, CI 95%: 2.75–15.72), and ethnicity (none) (OR: 2.13, CI 95%: 0.98–4.63). The latter did not reach statistical significance (see Table 3). Confounding and interaction variables were assessed, and none were identified.

Table 3

Sociodemographic factors associated with cardiovascular risk (Framingham*)

* Output/Framingham Variable: Age, total cholesterol, HDL, systolic blood pressure, smoking, and diabetes mellitus; **Multivariate odds ratios adjusted for other variables.

Discussion

This study's results suggest that there is a statistically significant difference in cardiovascular risk levels among patients enrolled in a chronic condition care program in the department of Putumayo, depending on their geographic location. This finding aligns with a 2013 study¹⁷ conducted in an urban population, which reported high CVR in 98.7% of the study population. However, it contrasts with the rural population findings of that study, where only 2.3% exhibited high CVR, a numerical difference compared to the 83.61% identified in this study. The high proportion of cardiovascular risk observed in Putumayo may be attributed to the fact that individuals who visit these IPS health centers are patients with symptoms or a history of cardiovascular risk-related conditions, as opposed to healthy individuals who also access healthcare services.

We found that educational level, area, and type of health affiliation were associated with higher cardiovascular risk. In the final logistic regression model, educational level was statistically significant in the technical/university/other category, indicating that the possibility of having a higher level of education among patients with high cardiovascular risk is one-third that of patients who are not at high cardiovascular risk. This finding is consistent with a study published in April 2019, which related low educational levels to higher cardiovascular mortality and suggested that individuals with higher education are more knowledgeable about the disease, its risk factors, and preventive measures⁵,¹⁸.

Subsidized health affiliation was identified as a risk factor for developing severe cardiovascular problems compared to contributory health affiliation. As observed in previous studies, no statistically or clinically significant association was found between cardiovascular risk and type of health affiliation¹⁹. However, according to a recent study, individuals with a subsidized health affiliation exhibited better glycemic index control than those with a contributory health affiliation²⁰. Given that diabetes and hypertension contribute to or increase CVR, glycemic control can be considered an indirect indicator of cardiovascular risk. According to the healthcare system's functioning, high cardiovascular risk levels are expected to be distributed independently of the type of health affiliation. However, in this study sample, the high proportion of individuals with subsidized health affiliation may indicate a possible association. This finding could also be explained by selection bias.

Ethnicity in the final logistic regression model was not statistically significant, with an OR of 2.13 and a 95% CI of 0.98–4.63. However, the literature shows that ethnicity is a predisposing factor for cardiovascular disease. A document published by the Spanish Heart Society cites multiple studies identifying a greater predisposition to AHT among individuals of black ethnicity, with evidence suggesting a worse prognosis in this population²¹,²². However, in Colombia, there is a mixture of cultures and races; therefore, these characteristics are measured by official entities using patient self-reporting, which poses challenges at the moment of objectively classifying them for data analysis. Additionally, for this analysis, the ethnicity variable was analyzed by grouping individuals at risk —Romanies, Gypsies, Black, and Afro-Colombian populations—compared to White individuals.

This study made it possible to know cardiovascular risk levels among patients attending chronic care units and living in rural and urban areas, revealing statistically significant differences between these areas in the department of Putumayo. These findings, in turn, provide the basis for formulating hypotheses in future research, as the study offers an overview of potential relationships between the variables of interest.

It is important to mention that the main limitation of this study is the apparent selection bias, as the sample consisted of patients enrolled in chronic care programs. This is why these individuals are more likely to have risk factors, which could lead to overestimating cardiovascular risk. It is relevant to recognize that not all patients with risk factors fall into the high-risk category. Therefore, conducting risk classifications during patient follow-up is important for improving the accuracy of cardiovascular risk assessment, particularly for subclinical cardiovascular disease, silent ischemia, and future cardiovascular events. Additionally, risk classifications could prevent the unnecessary use of treatments in low-risk patients, thereby increasing adverse events²³. It should also be mentioned that these programs' recruitment rates of chronic patients do not reach the entire population with CVR. However, an attempt was made to mitigate this bias by selecting participants through simple random sampling.

he possibility of measurement bias cannot be ruled out, as the main source of information was medical records, which in these municipalities are transcribed manually. Additionally, misclassification bias may have occurred, as errors in the variables used in the equation for calculating cardiovascular risk could have led to a change in the categorization of patients (low, moderate, or high risk). Consequently, such misclassification of information could also impact the interpretation of the results.

The importance of cardiovascular risk assessment and the identification of risk factors has long been a subject of research. Although many risk-enhancing factors have been identified, it is necessary to determine whether this risk is the same in different populations. Current health patterns, urbanization, dietary changes, access to healthcare services, and medical care may also influence the development of cardiovascular disease. This will enable the identification of potential socioeconomic factors to be considered in populations with specific characteristics, allowing the implementation of long-term, sustainable strategies to improve health outcomes. In turn, this could help mitigate the rising prevalence of cardiovascular diseases and reduce health costs¹⁸.

Implications for practice

The implications for clinical practice of this study lie in its potential to enhance the accuracy of cardiovascular risk classification and, consequently, optimize medical care. The findings of this study reveal significant differences in cardiovascular risk between rural and urban areas of Putumayo, highlighting the importance of adapting prevention and treatment strategies to the specific characteristics of each population. In clinical practice, these insights would allow health professionals to implement more personalized interventions, tailored to patients' socioeconomic and geographic conditions. Ultimately, this study establishes a starting point for future research and changes in clinical practice that will allow more efficient management of cardiovascular diseases, contributing to better public health and reducing the associated healthcare costs.

Conclusions

This study determined that there is a high proportion of individuals with elevated cardiovascular risk in Putumayo, with statistically significant differences between rural and urban areas. We declare that, due to selection bias, the population with which the statistical analyses were performed is not representative of the entire population of Putumayo. Therefore, the data can only be extrapolated to populations with chronic conditions. Cardiovascular risk was associated with educational level, ethnicity, type of health affiliation, and area.

Acknowledgment:

The Fundación Universitaria de Ciencias de la Salud and the Universidad Cooperativa de Colombia for providing the information on the IPS databases.

References

Almeida-Filho N, Castiel LD, Ayres JR. Risk: basic concept of epidemiology. Salud Colect. 2009;5(3):323-44. https://doi.org/10.18294/sc.2009.238

Cardiavant. Riesgo Cardiovascular 2024 Consulta: Abril 24, 2024. Disponible en: https://cardiavant.com/riesgo-cardiovascular/

Lopez-Jaramillo P, Lopez-Lopez J, Cohen D, Alarcon-Ariza N, Mogollon-Zehr M. Epidemiology of Hypertension and Diabetes Mellitus in Latin America. Curr Hypertens Rev. 2021;17(2):112-20. https://doi.org/10.2174/1573402116999200917152952

Ministerio de Salud y Protección Social. Situación de la enfermedad renal crónica, la hipertensión arterial y diabetes mellitus en Colombia 2020 - Cuenta de Alto Costo Consulta: Abril 24, 2024. Disponible en: https://cuentadealtocosto.org/publicaciones/situacion-de-la-enfermedad-renal-cronica-la-hipertension-arterial-y-diabetes-mellitus-en-colombia-2020/

Iglesias S. Un nuevo estudio relaciona el bajo nivel educativo con mayor mortalidad cardiovascular. Sociedad Española de Cardiología. 2019 Consulta: Abril 16, 2024. Disponible en: https://secardiologia.es/comunicacion/notas-de-prensa/notas-de-prensa-sec/10482-un-nuevo-estudio-relaciona-el-bajo-nivel-educativo-con-mayor-mortalidad-cardiovascular

Schultz WM, Kelli HM, Lisko JC, Varghese T, Shen J, Sandesara P, et al. AHA Circulation. 2018; 137 (20) 2166-2178. Socioeconomic Status and Cardiovascular Outcomes. Circulation. 2018;137(20):2166-78. https://doi.org/10.1161/CIRCULATIONAHA.117.029652

Szlenk-Czyczerska E, Kurpas D. Indicators of integrated care for patients with chronic cardiovascular disease in ambulatory care. Adv Clin Exp Med Off Organ Wroclaw Med Univ. 2023;32(10):1159-66. https://doi.org/10.17219/acem/161462

OECD. Health at a glance 2023 [Internet].2023. [cited 2024 Apr 17].Available from: https://www.oecd-ilibrary.org/social-issues-migration-health/health-at-a-glance-2023_7a7afb35-en

Organización Mundial de la Salud. Enfermedades no transmisibles. 2023. Consulta: Abril 17, 2024]. Disponible en: https://www.who.int/es/news-room/fact-sheets/detail/noncommunicable-diseases

Herrán OF, Patiño GA, Castillo SED. Dietary transition and excess weight in adults according to the Encuesta de la Situación Nutricional en Colombia, 2010. Biomédica. 2016;36(1):109-20. https://doi.org/10.7705/biomedica.v36i1.2579

Zambrano Bermeo RN, Rubiano Mesa YL. Enfermedad cardiovascular en mujeres rurales: una revisión sistemática integrativa. En: Zambrano Bermeo RN y Marín Muñoz JA (eds. científicos). Mujeres, Salud y Cotidianidad. ¿Iguales o diferentes? Cali, Colombia: Editorial Universidad Santiago de Cali; 2020. p 15-37. https://libros.usc.edu.co/index.php/usc/catalog/book/196

Álvarez-Ceballos JC, Álvarez-Múñoz AM, Carvajal-Gutiérrez W, González MM, Duque JL, Nieto-Cárdenas OA. Determinación del riesgo cardiovascular en una población. Rev Colomb Cardiol. 2017;24(4):334-41. https://rccardiologia.com/previos/RCC%202017%20Vol.%2024/RCC_2017_24_4_JUL-AGO/RCC_2017_24_4_334-341.pdf

David W. Hosmer Jr., Stanley Lemeshow, Rodney X. Sturdivant Applied Logistic Regression [Internet]. 2013. [citado 3 de junio de 2024]. https://doi.org/10.1002/9781118548387

Ruiz-Sandoval JP, Vizcaino-Sulbarán D, Álzate-Granados JP, Cáceres-Rivera DI, López-Romero LA. “Factores asociados con alto riesgo cardiovascular en el Putumayo.”, Mendeley Data, V1, 2024. https://doi.org/10.17632/gf5m2cjjzp.1

Abajo FJ. La Declaración de Helsinki VI: una revisión necesaria, pero ¿suficiente? Rev Esp Salud Pública. 2001;75(5):407-20. https://scielo.isciii.es/scielo.php?script=sci_arttext&pid=S1135-57272001000500002

Ministerio de salud. Resolución 8430 de 1993. Consulta: Noviembre 7, 2023. Disponible en: https://www.redjurista.com/Documents/resolucion_8430_de_1993.aspx

Navas JA. Estimación del riesgo cardiovascular con la escala Framingham calibrada, en población santandereana. [Tesis Especialista en Medicina Interna] Floridablanca: Facultad de medicina Departamento de medicina interna, Universidad Autónoma de Bucaramanga; 2018 https://repository.unab.edu.co/handle/20.500.12749/1732

Teo KK, Rafiq T. Cardiovascular Risk Factors and Prevention: A Perspective From Developing Countries. Can J Cardiol. 2021;37(5):733-43. https://doi.org/10.1016/j.cjca.2021.02.009

Vélez-Álvarez C, Gil-Obando L, Ávila-Rendón C, López-López A. Factores de riesgo cardiovascular y variables asociadas en personas de 20 a 79 años en Manizales, Colombia. Universidad y Salud 2015;17(1):32-46. http://www.scielo.org.co/scielo.php?script=sci_arttext&pid=S0124-71072015000100004

Urina-Jassir M, Herrera-Parra LJ, Hernández Vargas JA, Valbuena-García AM, Acuña-Merchán L, Urina-Triana M. The effect of comorbidities on glycemic control among Colombian adults with diabetes mellitus: a longitudinal approach with real-world data. BMC Endocr Disord. 2021;21(1):128. https://doi.org/10.1186/s12902-021-00791-w

Fundación Española del Corazón. Raza - Etnia. Consulta: Mayo 13, 2024. Disponible en: https://fundaciondelcorazon.com/prevencion/marcadores-de-riesgo/raza-etnia-linaje.html

Rojas J, Bermúdez V, Leal E, Aparicio D, Peña G, Acosta L, et al. Origen étnico y enfermedad cardiovascular. Arch Venez Farmacol Ter. 2008;27(1):40-57. http://ve.scielo.org/scielo.php?script=sci_abstract&pid=S0798-02642008000100009&lng=es&nrm=iso&tlng=es

Ribeiro LK, Amorim WW, Cardoso ITA, Vieira WS, Kochergin CN, Medeiros DS de, et al. Comparison of cardiovascular risk calculators in patients with diabetes. Rev Assoc Médica Bras. 2021;67:200-6. https://doi.org/10.1590/1806-9282.67.02.20200514

Notas

Autor notes

^*Correspondence: Jessica Paola Ruiz-Sandova.jrsandoval9618@gmail.com