Comparative Analysis of Cardiovascular Disease risk in a local population of Sri Lankans using risk predictors: WHO ISH charts, Original and recalibrated Framingham Equation

doi:10.21203/rs.3.rs-4997260/v1

Background: Cardiovascular diseases (CVDs) accounted for 32% of global deaths in 2019, mostly in lower and middle-income countries and ischemic heart diseases and stroke were leading causes. Prevention relies on individual CVD risk prediction using methods like the Framingham Risk Equation, SCORE, and WHO ISH charts for effective screening and management.This study focuses on assessing and comparing CVD risk using different WHO ISH charts and both traditional and recalibrated Framingham equations and calculating the agreement between specific risk predictors.

Methods:CVD risk was assessed using WHO ISH Chart with cholesterol levels of 5mmol/dl and 4mmol/dl, and individual cholesterol levels. Additionally, the non-cholesterol WHO ISH Chart was used. The four values were compared using the Cohen Kappa statistic. CVD risk was calculated using the traditional Framingham Equation and a recalibrated version with local data. These were compared with WHO ISH Chart results using Cohen Kappa statistic. The Framingham scores were categorized per WHO ISH categories. Multivariate analysis was performed using SPSS 25 to find out the association between risk factors not used for risk calculation with the predicted cardiovascular disease risk.

Results:WHO ISH methods indicated a 10-year CVD risk below 10% for over 90% of the population. Original and recalibrated Framingham scores predicted 55.5% and 62.3% with less than 10% risk, respectively. Significant gender differences were noted in high CVD risk category between the original Framingham (13.8% females vs. 21.3% males) and recalibrated Framingham (9% females vs. 13.4% males) risk equation. WHO ISH cholesterol method showed substantial agreement (k=0.716, p<0.001) with WHO ISH assume 5 method, while the non-cholesterol method had slight agreement (k=0.124, p<0.01) with WHO ISH assume 4 method. Substantial agreement (k=0.613, p<0.01) was found between original and recalibrated Framingham scores. The original Framingham score had poor agreement with WHO ISH methods, but recalibration improved the values, with the highest being fair agreement (k=0.365, p<0.01) with the WHO ISH cholesterol method.

Conclusion:Different CVD risk prediction methods yielded varying 10-year risk distributions. Significant differences were observed between original and recalibrated Framingham scores. Varying levels of agreement were observed between risk predictors. Associations were also found between non-calculated risk factors and predicted CVD risk.

CVD

Risk prediction

Sri Lanka

Original Framingham scores

Recalibrated Framingham scores

WHO ISH Chart

age group

sex

agreement

none calculated variables

Cardiovascular diseases (CVDs), one of the top 10 causes of death according to WHO (WHO, 2020b) accounts for very high rates of mortality with 75% occurring in lower and middle-income countries(WHO, 2021). Due to the grievous impact associated with the individuals and the society, preventing CVDs is imperative.

For this approach, it is essential to screen and manage the risk factors, according to the predicted CVD risk of an individual(WHO, 2007). Predicting the CVD risk is conducted through various CVD risk prediction methods, incorporating a variable number of risk factors. There are many CVD risk prediction methods developed and used throughout the world. Some of them include: Framingham Risk Equation, Systematic Coronary Risk Evaluation (SCORE) Risk Charts and World Health Organization International Society of Hypertension (WHO ISH) CVD risk prediction charts.

Both Framingham Risk Equation and SCORE Risk Charts were developed using findings of cohort studies. WHO ISH chart has been developed using relative risks assessed in risk level assessment projects combined with absolute risks from the global burden of disease studies .The internal and external validity of WHO ISH Chart is not studied (Cooney et al., 2009). Both Framingham Risk Equation and SCORE Methods have high internal validity for discrimination and a high external validity (Aktas et al., 2004, Scheltens et al., 2008, Cooney et al., 2010). A Malaysian study has shown that WHO/ISH Model performed poorly in stratification of CVD Risk than both Framingham Risk Equation and SCORE Models for CVD Risk Prediction (Selvarajah et al., 2014).

WHO ISH CVD Risks chart does the calculation using selected factors; Diabetes mellitus (DM) status, sex, smoking and Blood Pressure. Currently there are many other factors known to cause CVD such as Body Mass Index (BMI), family history of CVD, physical activity levels, dietary patterns and psychosocial index.

In Sri Lanka there has been an elevation of case fatality rates of CVD from 24.92% in 2004 to 28.05% in 2019 with 398027 case load reported in 2019( Senaviratna, 2023). Moreover, the ministry of health has been using the WHO ISH chart with cholesterol for CVD risk screening even at settings without total cholesterol measurement facilities. There is a separate chart developed by the WHO to be used at such poor resource settings(WHO, 2020a) and these two charts have not been compared previously regarding assumed total cholesterol levels which is a novel objective in this study. This study compares and assesses the risk of CVD according to age groups and gender using several WHO ISH charts, traditional and recalibrated Framingham equations, and measures the degree of agreement between separate risk predictors. In addition the study focuses on the association between risk factors that are not used in risk calculation by CVD risk predictors.

Sample selection and Data collection

This study is an elaborated analysis of the data collected from the work of Ranasinghe et al which was a cross-sectional study conducted in the Kurunegala Regional Director of Health Services area where residents between 30- to 59-years-old who has been living within the area for a period of more than one year were selected. The following formula N=\(\:{\left(\text{Z}{\alpha\:}\right)}^{2}\text{*}\text{P}\left(1\:-\:\text{P}\right)\text{*}\text{D}\text{E}/{d}^{2}\) (Lemeshow, 1991) was used to calculated the sample size (N= 1102) where Zα and P were level of statistical significant (3.92) and the expected high CVD risk proportion (7%) at 0.025 precision and design effect to be 2 respectively. Participants were selected from the study population using multi-stage cluster sampling excluding the individuals who either suffering from CVD endpoints or already followed by healthy life style centres. Data with regard to demographic features (age, sex, family history of CVDs), anthropometric features (BMI, blood pressure), biochemical features (Lipid profile, Fasting Blood Sugar) and lifestyle factors (physical activity level, daily vegetables and fruit consumption) were collected from eligible participants. The data collected for the above study was used to carry out the CVD risk assessment using WHO ISH charts and both traditional and recalibrated Framingham Equations. Ethical approval was obtained from the Ethical review committee at faculty of Medicine, University of Peradeniya and prior to participation, informed written consent was taken from all the participants.

CVD risk assessment

First, CVD risk was measured using the WHO ISH Chart assuming a mean total Cholesterol level of 5 mmol/dl (used by the Health Ministry of Sri Lanka for WHO ISH charts) and 4 mmol/dl (mean total Cholesterol level of Sri Lankans according to STEPS survey was 4mmol/dl). Then the individual total Cholesterol levels were used and again the WHO ISH Risk level was assessed. Furthermore, WHO ISH non-cholesterol chart was used to assess the risk. (Annex I). These four values were compared using Cohen Kappa statistic.

Secondly, the risk score was calculated using the traditional Framingham Equation, derived from the Framingham Cohort Study. Then the Framingham Equation was recalibrated using the local risk factor and morbidity and mortality level data (Ranasinghe et al., 2024). CVD risk was calculated with the recalibrated equation which was compared with the CVD risk values obtained with the original Framingham Formula. The two Framingham Formulae CVD Scores were categorized in accordance with WHO ISH categories. Thirdly, the four CVD risks calculated from WHO ISH Charts and the two CVD risks calculated from the Framingham Equation were compared using Cohen Kappa statistic. Data analysis was carried out using the SPSS 25 software.

Cardiovascular Disease risk predicted by different prediction methods

By all the four WHO ISH based methods, majority of the population (more than 90%) were found to have a predicted 10-year CVD risk below 10%. When using the WHO ISH Cholesterol method 5.1% (N = 56) and 2.5% (N = 28) of the population were found to have 10-year CVD risk between 10%-20% and 20–30% respectively. In the above method, 0.8% (N = 9) were found to have a 10-year CVD risk over 40%. With the WHO ISH assume 5 and WHO ISH assume 4, the percentages of the sample having CVD risk more than 10% were, 3.3% and 1% respectively. WHO ISH non cholesterol method predicted that only 1.7% were having CVD risk of over 10% (Table 01).

Table 01

Percentage distribution of predicted cardiovascular diseases risk categories by selected risk estimate methods
10-year CVD risk	WHO ISH without cholesterol N (%)	WHO ISH with individual cholesterol N (%)	WHO ISH with assumption of total cholesterol 5mmol/dl	WHO ISH with assumption of total cholesterol 4mmol/dl		Recalibrated Framingham CVD score
< 10%	1083(98.3)	1002(90.9)	1066(96.7)	1091(99)	612(55.5)	687(62.3)
>=10%-<20%	11(1.0)	56(5.1)	11(1.0)	8(0.7)	303(27.5)	295(26.8)
>=20%-<30%	5(0.5)	28(2.5)	22(2.0)	3(0.3)	124(11.3)	76(6.9)
>=30%-<40%	2(0.2)	90(0.8)	3(0.3)	0(0.0)	39(3.5)	24(2.2)
>=40%	1(0.1)	7(0.6)	0(0.0)	0(0.0)	24(2.2)	20(1.8)

Original and recalibrated Framingham CVD risk scores predicted 55.5% (N = 612) and 62.3% (N = 687) to be having less than 10% CVD risk respectively. Further, the original and recalibrated CVD Risk Scores predicted 2.2% (N = 24) and 1.8% (N = 20) to be having CVD risk more than 40% respectively (Table 01).

Since the numbers in higher risk categories were low, cardiovascular Diseases risk predicted by different CVD scores were categorized into two categories as CVD risk less than 20% and CVD risk greater than or equal to 20% for the purpose of the following comparisons.

Distribution of cardiovascular diseases risk by sex

Cardiovascular Diseases risk predicted by different CVD scores were categorized into two categories, CVD risk less than 20% and CVD risk greater than or equal to 20%. Compared to other methods highest percentage of females were marked as having > 20% CVD risk by the original Framingham Score (N = 87, 13.8%) followed by recalibrated Framingham Score (N = 57, 9%). A similar pattern was observerd among males, where 21.3% % (N = 100) were categorized in to high CVD risk by original Framingham Scores followed by recalibrated Framingham Score (N = 63, 13.4%). There were significant differences by sex in categorization in to > 20% CVD risk in recalibrated Framingham Method (9% females vs. 13.4% males, p = 0.02) and original Framingham Method (13.8% females vs. 21.3%% males, p = 0.01) (Table 2).

Table 2

Distribution of CVD risk by different CVD risk scores according to sex
		Female	Male	P Value
		N (%)	N (%)	P Value
WHO ISH without cholesterol	> 20%risk	2(0.3)	6(1.3)	0.063
	< 20%risk	630(99.7)	464(98.7)
	Total	632(100.0)	470(100.0)
WHO ISH with individual cholesterol	> 20%risk	34(5.4)	10(2.1)	0.04
	< 20%risk	605(94.6)	453(97.9)
	Total	632(100.0)	470(100.0)
WHO ISH with assumption of total cholesterol 5mmol/dl	> 20%risk	16(2.5)	9(1.9)	0.49
	< 20%risk	616(97.5)	461(98.1)
	Total	632(100.0)	470(100.0)
WHO ISH with assumption of total cholesterol 4mmol/dl	> 20%risk	1(0.2)	2(0.4)	0.4
	< 20%risk	631(99.8)	468(99.6)
	Total	632(100.0)	470(100.0)
Original Framingham score	> 20%risk	87(13.8)	100(21.3)	0.01
	< 20risk	545(86.2)	370(78.7)
	Total	632(100.0)	470(100.0)
Recalibrated Framingham score	> 20%risk	57(9.0)	63(13.4)	0.021
	< 20%risk	575(91.0)	407(86.6)
	Total	632(100.0)	470(100.0)

Distribution of cardiovascular risk by age group

Framingham Original CVD Risk Score categorized the highest percentages of individuals (15.2%, N = 56, 34.9%, N = 129) into high CVD risk under the 40-49- and 50-59-years age groups compared to other methods. Recalibrated Framingham CVD risk score categorized the highest percentage of individuals (1.9%, N = 7) into high CVD risk in the 30–39 years age group. WHO ISH Cholesterol Method categorized 1.7% (N = 6), 4.3% (N = 16) and 5.9% (N = 22) of the sample into high CVD risk categories in 30–39, 40-49- and 50-59-years age groups respectively. In every age category lowest percentage of the participants (0.0%, 0.3% and 0.5% under 30–39, 40-49- and 50-59-years age groups respectively) were categorized into high CVD Risk by WHO ISH assume 4. CVD risk category (high and low) and age groups were significantly associated with Framingham methods and WHO ISH cholesterol method (Table 3).

Table 3

Distribution of cardiovascular diseases risk when using different cardiovascular diseases risk scores according to Age Groups
		Age Category			P value
		30–39	40–49	50–59
		N (%)	N (%)	N (%)
WHO ISH without cholesterol	> 20%risk	2(0.6)	1(0.3)	5(1.4)	0.20
	< 20%risk	361 (99.4)	368(99.7)	365(98.6)
	Total	363(100.0)	369 (100.0)	370(100.0)
WHO ISH with individual cholesterol	> 20%risk	6(1.7)	16(4.3)	22(5.9)	0.011
	< 20%risk	357(98.3)	353(95.7)	348(94.1)
	Total	363 (100.0)	369(100.0)	370(100.0)

WHO ISH with assumption of total cholesterol 5mmol/dl	>20%risk	5(1.4)	12(3.3)	8(2.2)	0.23
	<20%risk	358(98.6)	357(96.7)	362(97.8)
	Total	363(100.0)	369(100.0)	370(100.0)
WHO ISH with assumption of total cholesterol 4mmol/dl	>20%risk	0(0.0)	1(0.3)	2(0.5)	0.37
	<20%risk	363(100.0)	368(99.7)	368(99.5)
	Total	363 (100.0)	369(100.0)	370(100.0)
Original Framingham score	>20%risk	2(0.6)	56(15.2)	129(34.9)	<0.01
	<20risk	361 (99.4)	313(84.8)	241(65.1)
	Total	363(100.0	369(100.0)	370(100.0)
Recalibrated Framingham score	>20%risk	7(1.9)	18(4.9)	95(25.7)	<0.01
	<20%risk	356(98.1)	351(95.1)	275(74.3)
	Total	363(100.0)	369(100.0)	370(100.0)

Agreement between WHO ISH cholesterol method and WHO ISH non –cholesterol method

The highest percentage (56.8%) of people who were categorized as high CVD risk by the WHO ISH cholesterol method were categorized into high CVD risk by WHO ISH assume 5 when compared to other methods. It was found that the agreements between the WHO ISH Cholesterol method had a “substantial agreement” (k = 0.716, p < 0.001) between the WHO ISH assume 5 method. WHO ISH Cholesterol method had a “slight agreement” (k = 0.124, p < 0.01) between the WHO ISH assume 4 method and WHO ISH non cholesterol method (k = 0.104, p < 0.01) (Table 4).

Table 4

Percentage agreement between WHO ISH cholesterol method and WHO ISH non-cholesterol methods
		WHO ISH cholesterol		Kappa
		> 20%risk	< 20%risk
		N (%)	N (%)
WHO ISH without cholesterol	> 20%risk	3(6.8)	5(0.5)	K = 0.104, SE = 0.06, Tb = 4.56, p < 0.01
	< 20%risk	41(93.2)	1053(99.5)
	Total	44(100.0)	1058(100.0)
WHO ISH with assumption of total cholesterol 5mmol/dl	> 20%risk	25(56.8)	0(0.0)	K = 0.716, SE = 0.062, Tb = 24.801, p < 0.01
	< 20%risk	19(43.2)	1058(100.0)
	Total	44(100.0)	1058(100.0)
WHO ISH with assumption of total cholesterol 4mmol/dl	> 20%risk	3(6.8)	0(0.0)	K = 0.123, SE = 0.065, Tb = 8.505, p < 0.01
	< 20%risk	41(93.2)	1058(100.0)
	Total	44(100.0)	1058(100.0)

Agreement between Framingham Original Cardiovascular Disease Score and recalibrated Framingham Cardiovascular Disease Score

Original framingham score categorized 97 individuals out of the 120 (80.8%) that were religated as high risk by the recalibrated score as individuals with high CVD risk. It was found that the agreements between the Original Framingham CVD Risk Score and Recalibrated Framingham CVD Risk Scores was a “substantial agreement” (k = 0.613, p < 0.01) (Table 5).

Table 5

Agreements between Framingham Original Cardiovascular Disease Score and Recalibrated Framingham Cardiovascular Disease Score
		Recalibrated Framingham Score		Kappa
		> 20%risk	< 20%risk
		N (%)	N (%)
Original Framingham Score	> 20%risk	97 (80.8)	90(9.2)	K = 0.613, SE = 0.034, Tb = 21.03, p < 0.01
	< 20risk	23(19.2)	892(90.8)
	Total	120(100.0)	982(100.0)

Agreement between Framingham original cardiovascular disease Score and WHO ISH cardiovascular disease score

Out of WHO ISH CVD risk methods percentage agreements with original Framingham Score in the high-risk category was highest (5.3%) in WHO ISH cholesterol method. All WHO ISH methods had poor agreement with original Framingham CVD risk score (Table 6).

Table 6: Agreements between the Framingham Original Cardiovascular Diseases Score and WHO ISH Cardiovascular Diseases Scores

		Original Framingham Score		Kappa value
		>20%risk	<20risk
		N (%)	N (%)
WHO ISH with cholesterol	>20%risk	10(5.3)	34(3.7)	K=0.023, SE=0.025, Tb=1.04, p<0.05
	<20%risk	177(94.7)	881(96.3)
	Total	187(100.0)	915(100.0)

WHO ISH without cholesterol	>20%risk	4(2.1)	4(0.4)	K=0.03, SE=0.017, Tb=2.49, p<0.05
	<20%risk	183(97.9)	911(99.6)
	Total	187(100.0)	915(100.0)

WHO ISH with assumption of total cholesterol 5mmol/dl	>20%risk	6(3.2)	19(2.1)	K=0.02, SE=0.021, Tb=0.947, p=0.34
	<20%risk	181(96.8)	896(97.9)
	Total	187(100.0)	915(100.0)

WHO ISH with assumption of total cholesterol 4mmol/dl	>20%risk	1(0.5)	2(0.2)	K=0.01, SE=0.01, Tb=0.76, p=0.45
	<20%risk	186(99.5)	913(99.8)
	Total	187(100.0)	915(100.0)

Agreement between recalibrated Framingham cardiovascular disease risk score and WHO ISH CVD risk score

In high CVD risk category highest agreement (27.5%) was between the WHO ISH cholesterol method and recalibrated Framingham score. In low CVD risk category highest agreement (100%) was with the WHO ISH with assume 5 methods. After recalibration, Framingham CVD Risk Score showed improvements in agreements between the WHO ISH CVD Risk Scores. The improvement was highest with WHO ISH Cholesterol method, showing “fair agreement” (k = 0.365, p = < 0.01). A moderate improvement was seen in the agreement with WHO ISH assume 5 methods, showing “fair agreement” (k = 0.233, p = < 0.01). The agreement of Recalibrated Framingham CVD Risk Score with the WHO ISH Non-Cholesterol method and WHO ISH assume 4 were poor (k = 0.05, p < 0.01) and (k = 0.027, p < 0.01 respectively) (Table 7).

Table 7: Percentage agreements between recalibrated Framingham cardiovascular disease risk score and WHO ISH CVD risk scores

		Recalibrated Framingham Score		Kappa value
		>20%risk	<20%risk
		N (%)	N (%)
WHO ISH with cholesterol	>20%risk	33(27.5)	11(1.10)	K=0.05, SE=0.027, Tb=3.564, p<0.01
	<20%risk	87(72.5)	971(98.9)
	Total	120(100.0)	982(100.0)

WHO ISH without cholesterol	>20%risk	4(3.3)	4(0.4)	K=.365, SE=.049, Tb=13.933, p<0.01
	<20%risk	116(96.7)	978(99.6)
	Total	120(100.0)	982(100.0)

WHO ISH with assumption of total cholesterol 5mmol/dl	>20%risk	25(20.8)	0(0.0)	K=0.233, SE=0.046, Tb=10.572, p<0.01
	<20%risk	95(79.2)	982(100.0)
	Total	120(100.0)	982(100.0)

WHO ISH with assumption of total cholesterol 4mmol/dl	>20%risk	2(1.7)	1(0.1)	K=0.027, SE=0.020, Tb=3.106, p<0.01
	<20%risk	118(98.3)	981(99.9)
	Total	120(100.0)	982(100.0)

Association between Risk Factors not used for risk calculation with the predicted cardiovascular disease risk

Association between co-morbidity and family history related risk factors, behavioural risk factors, anthropometric risk factors and biochemical risk factors with the CVD risk predicted by all CVD risk prediction methods were explored using multivariate analysis. World Health organization cholesterol (WHO-chole) CVD risk was associated significantly with antihypertensive medication (aOR = 1.97, CI = 1.05–3.68), inadequate daily fruit consumption (aOR = 3.09, CI = 1.42–6.7) and high Triglycerides (aOR = 1.83, CI = 1.03–3.42). Original Framingham Risk score ( FRS-Org) CVD risk was statistically significantly associated with inadequate fruit consumption (aOR = 1.578, CI = 1.22–1.94) and high Triglyceride levels (aOR = 1.709, CI = 1.217–2.399). High Triglyceride levels (aOR = 1.51 CI = 1.01–2.24), inadequate vegetable consumption (aOR = 1.697, CI = 1.491–1.91), high waist circumference (aOR = 1.355, CI = 1.149–1.559) were significantly predicting recalibrated Framingham Risk score (FRS-recal) CVD risk (Table 8).

Table 8

Association between Risk factors not used for risk calculation with the predicted cardiovascular disease risk
	WHO ISH Individual cholesterol					Original Framingham CVD risk equation					Recalibrated Framingham CVD risk equation
		95% Confidence Interval(C.I) for aOR*				(aOR*)	95% C.I.for aOR*				(aOR*)	95% C.I. for aOR*
	(aOR*)		Lower		Upper	(aOR*)	Lower			Upper	(aOR*)	Lower	Upper
On antihypertensive medication	1.97			1.05	3.68
Daily inadequate fruit consumption	3.09			1.42	6.7	1.578		1.22	1.941		1.181	0.547	1.302
Daily inadequate vegetable consumption	1.37			0.6	3.11	1.11		0.623	1.32		1.697	1.491	1.91
Low physical activity level
Moderate physical activity level	0.63			0.28	1.44	1.269		0.877	1.835		1.27	0.812	1.984
High physical activity level	2.74			0.29	0.37	1.484		0.907	2.427		1.184	0.658	2.129
High waist circumference	1.64			0.84	3.22	1.275		0.903	1.799		1.355	1.149	1.559
High Triglyceride level	1.83			1.03	3.42	1.709		1.217	2.399		1.503	1.007	2.243
Constant	0.01					3.874					5.366
*aOR-Adjusted Odds Ratio

Cardiovascular Disease Risk predicted by WHO ISH charts

In the present study, prevalence of High CVD Risk predicted by WHO ISH Cholesterol Method was 4% whereas none of the WHO ISH methods that does not use direct cholesterol measurements predicted high CVD risk above 3%. Study by Ranawaka at el. reported the prevalence of high CVD Risk proportion to be 8.2% when using non specified WHO ISH CVD Risk Charts (Ranawaka et al., 2016). In the study by Mettananda et al, prevalence of high CVD Risk was 10.7% and 9.5% when using WHO ISH Cholesterol and WHO ISH Non-Cholesterol methods respectively. The reason for the differences can be the age of the participants ( older participants )in both Mettananda et al and Ranawaka at el studies (Ranawaka et al., 2016; Mettananda et al., 2021). Additionally, present study includes 30–35 years age category which generally has a low CVD Risk due to the age( Dhingra and Vasan, 2012). This might have resulted in low prevalence of high CVD Risk in the present study. This may also explain the low proportion of high CVD Risk when using WHO ISH Charts compared to Mettananda et al. study. A cohort study in Sri Lanka by Thulani et al. has shown high percentages of high CVD Risk predictions compared to the present study by both WHO ISH Cholesterol and Non-Cholesterol Methods Sri Lankans (Thulani et al., 2021). The older age group of 40–64 years in that cohort study might have produced these different findings. In the present study in WHO lSH assume 5 method, only 0.3% were identified as having CVD Risk > 30%. When comparing with the annual health statistics of Sri Lanka, only 0.7% of screened population from Kurunegala district had high CVD risk > 30% when using WHO lSH assume 5 method (Ministry of Health, 2019). This might be an evidence of a specific characteristics of a population of a specific area. All the other studies have been conducted in urban areas with different population characteristics compared to the mostly semi urban Kurunegala district.

Further, results of the present study are similar to a Cuban study (40–80 years) when comparing the WHO ISH Cholesterol and Non-Cholesterol methods and Peruvian study when comparing non specified WHO ISH Method (Nordet et al., 2013; Bazo-Alvarez et al., 2015). What these findings show is that even in different regions of the world, using different risk prediction charts similar results can be seen due to similarities that can occur in CVD Risk factor levels. On the other hand, these results were lower than findings from UK South Asian migrant, Indian and Omani studies (Bansal et al., 2015; Al-Lawati et al., 2013; Findlay et al., 2020). These studies do not mention clearly whether the WHO ISH Cholesterol or Non-Cholesterol methods were used. The different levels of CVD risk factors and different risk charts (By WHO region), diabetic population in Omani study, and sample with MI in Indian study might have resulted in these different findings.

Other important factor we have to consider in the present study is that, WHO ISH Charts are not developed to be used for 30 to 39-year age group, (WHO, 2007). For the 30-to-39-year group we used the chart developed for 40–49 years groups which has been recommended for the 35–39 years age group as per guidelines of the ministry of Health, Sri Lanka. This might have produced low percentages of high CVD Risk category participants. The results from the present study have shown this by the low prevalence of high CVD Risk Prediction among persons in the 30–39 years age category by all four WHO ISH Methods.

Cardiovascular Disease risk predicted by original Framingham Risk Scores

Original Framingham Risk Scores identified 17% as having high CVD risk > = 20%. This is extremely low when compared to Mettananda et al and Ranawaka at el. Studies from Sri Lanka (, Ranawaka et al., 2016; Mettananda et al., 2021). The reason can be attributed to the high HDL Cholesterol level found in this study compared to other two studies. HDL Cholesterol level directly affects the measurement of Framingham Original CVD Risk score. High HDL Cholesterol levels lower the CVD Risk calculated with the Framingham Risk Scores CVD Equation. Further, the age groups used in those studies were older compared to the present study. Approximately one third of the participants in this study were in the 30–39 years age group and none were above or equal to 60 years. This might have resulted in these low percentages of high CVD Risk by original version of the Framingham Risk Scores in the present study. Further, the percentage on antihypertensive in Mettananda et al study was higher than 27.8% that was found in the present study. The Ranawaka at el study reported 2.7% with previous CVD events who were included in the risk calculation. When comparing with studies from other parts of the world, the proportions with high CVD risk categories are low in the present study (Al-Lawati et al., 2013; Selvarajah et al., 2014; Bansal et al., 2015; Garg et al., 2017,). High hypercholesterolemia percentages and low mean HDL Cholesterol in these studies might have resulted in the said differences. High mean HDL Cholesterol level in the present study provides a large negative value on BiXi function. Accordingly, it results in a large reduction in CVD risk.

Cardiovascular Disease Risk predicted by recalibrated Framingham Cardiovascular Disease Scores

In the present study, Framingham recalibrated version predicted a total of 10.9% (9.0% females and 13.4% males) to be in high CVD Risk categories which was lower than the risk calculated by the original version (13.8% in females and 21.8% males). In an Australian Indigenous study, the recalibrated Framingham Risk Scores returned a higher risk of 19.6% and 22.9% compared to 8.9% and 15.4% with the original version for males and females respectively (Hua et al., 2017). In a Hong Kong study, recalibrated Framingham Risk Scores had predicted 36.1% males and 22.2% females to be in the high CVD Risk category which was more or less similar to the risk scores by the original version (Leung et al., 2018). In the Hong Kong study, the population was older (mean 65 years) which may have given rise to the higher high CVD risk proportions. Although the Australian Indigenous study had a similar lower age cut-off the upper end was 74 as against the 60 years of the current study which may be the reason for the higher risk category being larger. In interpreting these results we have to take into consideration that the recalibration was conducted using beta coefficients of the Framingham Original Study replacing only the local risk factor mean values and CVD survival rate data. Comparatively different mean values and survival data might be producing different findings. Especially, when compared to these studies the proportion smoking among females is zero in the present study. The smoking has a higher impact on the equation itself. Therefore, the different results might have been obtained. When calculating the BiMi function the high mean of HDL Cholesterol level in the present study might have also led to low CVD Risk levels in the final calculation in the present study (refer-methodology section).

Agreements between Cardiovascular Disease risk prediction methods

At kappa 0.716 agreement between WHO ISH Cholesterol and WHO ISH assume 5 methods was found which, was roughly similar to the findings of the Mettananda et al study and the cohort study by Thulani et al (Mettananda et al., 2021, Thulani et al., 2021). Furthermore the percentage agreements between the two scores are high (100% of > 20% CVD risk persons of WHO ISH assume-5 method were categorised as > 20% CVD risk by WHO ISH Cholesterol method). The 5mmol/dl is the recommended assumption to be used for the total Cholesterol in a low resource setting. Apart from assuming cholesterol as five, a further calculation assuming a total Cholesterol of 4mmol/dl (WHO ISH assume 4) which, was the mean total Cholesterol level reported in the Sri Lankan WHO STEPS Study was carried out (WHO, 2017). Both the, WHO ISH assume 4 and the WHO Non cholesterol method did not provide good agreement (k = 0.104 and 0.123 respectively) with the WHO ISH Cholesterol method. On the other hand, in the present study, mean total Cholesterol level was 209 mg/dl (~ 5mmol/dl) which may have resulted in good agreement between WHO ISH Cholesterol Method and WHO ISH assume 5. This supports the use of the value recommended by the ministry of health of Sri Lanka of 5mmol/dl with WHO ISH CVD Risk assessment in a low resource setting., This finding further supports the claim that the risk prediction models which require more laboratory data tend to be in good agreement with models which don’t use much lab data (Gaziano et al., 2013; Jones et al., 2015).

The agreements between the original and recalibrated Framingham versions were substantial (k = 0.613). When considering low CVD risk strata, percentage agreements between the scores were higher compared to high CVD risk categories. The reason might be the different mean values of risk factors and survival rates used in the recalibration than Framingham Original population. The agreement between the original Framingham Risk Score and the WHO ISH Cholesterol Method was seen as a kappa of 0.023. When considering the high CVD risk strata, percentage agreement between WHO ISH charts and Framingham Original Scores were extermely low (only 10 out of 187 high CVD risk persons by Orginal Framigham scores being identified as high CVD risk by WHO ISH colesterol method). This does not agree with the study of Mettananda et al. where agreement kappa was 0.306 (Mettananda et al., 2021). The reason might be the 30–39 years old group of participants in the current study producing much lower CVD risk predictions with the WHO ISH charts where 40–49 years’ subscale was used to determine the CVD Risk of this particular age group.

Other important finding is that after recalibration, the Framingham Risk Score method showed improvement in agreement (k = 0.365 and k = 0.233) with both WHO ISH Cholesterol Method and non-Cholesterol Method, WHO ISH assume 5 method. The agreements between the methods were higher in low CVD risk strata.This further highlights the importance of recalibrating Framingham CVD Risk Formulae using local risk factor levels and survival rates. Even though this study cannot provide evidence for validity of WHO ISH Cholesterol Methods, as it is the recommended and in use method in Sri Lanka it can be considered as the gold standard for risk estimation for Sri Lanka. Once recalibrated for local risk factor levels Framingham CVD Equation shows high agreement with the WHO ISH Cholesterol Method. The important fact that has to be taken into consideration while interpreting this finding is that the recalibration was conducted using the mean risk factors data from the local population. The original beta coefficients of the Framingham Study were not changed due to the unavailability of local beta coefficients. Locally applicable beta coefficients have to be developed through a cohort study in the local context. The improvement could be much higher if such developed coefficients were used for the recalibration. If that agreement was made higher, both the recalibrated Framingham version and the WHO ISH Chart may be used interchangeably. Since Framingham Equations take into consideration HDL Cholesterol levels and hypertension status into their risk prediction model when compared to WHO ISH Methods, the former can be considered superior to the latter. On the other hand, since WHO ISH CVD Charts does not depend much on laboratory methods for risk levels to predict CVD risk. It can be advantageous in economic terms in low resource settings. Further, the WHO ISH and a recalibrated Framingham method further improved by using local beta coefficients could be interchangeably used even for high-risk categories.

Age and sex wise distributions of predicted Cardiovascular Disease Risk

In the present study Original Framingham CVD Risk Equations predicted significantly higher proportion of males than females to have high CVD risk (21.3% vs. 13.8%). Similarly in Mettananda et al study and in an Iranian study Framingham Original CVD Risk Score predicted significantly high proportion of males in to high CVD risk category (Nakhaie et al., 2018; Mettananda et al., 2021). High dyslipidaemia levels among males in these studies might have resulted in these high CVD risk calculations (Tennakoon et al., 2010).

In this study, the CVD risk predicted by WHO ISH method has categorized a significantly high percentage of females into high CVD risk category compared to males (5.4% vs. 2.1%). This is similar to a study by Ranawaka et al. (Ranawaka et al., 2016) and a Nepali study which reported, CVD Risk categorization dependent on sex (Khanal et al., 2017). The Ranawaka et al study shows similar high proportion of females having high CVD risk whereas Nepali study shows the opposite. In a Jamaican study using WHO ISH Charts, there was no differences in CVD risk categorization according to sex (Tulloch-Reid et al., 2013). Nepal and Sri Lanka being south Asian countries tend to have similar CVD risk factor level prevalence compared to Jamaica which may have resulted in the findings.

CVD risk categorization was dependent on age in CVD risk predicted by WHO ISH cholesterol method and both Framingham methods in the present study, with prevalence in the high-risk category increasing with age. Similarly in Ranawaka et al study and three Asian country study CVD risk categorization was influenced by age (Otgontuya et al., 2013; Ranawaka et al., 2016). The noteworthy fact is that in 30–39 years age group there were fairly low percentage of high CVD risk persons when using any method in this study.

Associations of risk factors not included in the risk prediction models with the predicted risk

Risk prediction models include selected risk factors out of many cvd risk factors. Here we attempted to explore the associations between the risk predicted by different models with risk factors not included in the respective models.

In the present study, CVD risk predicted by any of the methods was not associated with the family history of CVD. In general, family history is a known and an important factor associated with future CVD (Schunkert et al., 2011). In Epic Norfok study, it was shown that Family history of CHD was associated with high CHD risk by Framingham equation (Sivapalaratnam et al., 2010). However,this study was conducted using the hospital records and only used CHD events in contrast to the present study (Sivapalaratnam et al., 2010). Family history was found to be a significant predictor of CVD risk assessed by the WHO ISH CVD risk charts (Non-Cholesterol Method) in multivariate analysis in Indian Kurnool area study where the association was derived through a proxy indicator e.g.; through direct CVD Risk factors like DM and hypertension to depict the predicted CVD risk. In the present study, family history of CVD events was recorded as reported by the participants which may have been inaccurate in some of the instances which may have given rise to the differences in the findings compared to other studies. Findings similar to our study were reported by a Malaysian study with regard to WHO ISH Cholesterol Method (Norhayati et al., 2013). However, neither WHO ISH Method nor Framingham Method incorporates family history in to their equations (Assmann et al., 2002, Woodward et al., 2007). Findings from the present study suggest that either of these risk predictions does not represent the risk imposed by family history of events.

In the present study, being on antihypertensive treatment was a significant factor associated with the CVD risk predicted by WHO ISH Cholesterol Method. Similarly, in a Punjabi study, it was shown that the high-risk category by WHO ISH Cholesterol was associated with use of antihypertensive. However, in the above-mentioned study, the bivariate or multivariate analysis was not conducted to determine the significance (Vikramaditya et al., 2017). WHO ISH Cholesterol Method does not take in to account the fact of being on antihypertensive for their calculations as the Framingham Equation does. However, from these study findings it can be assumed that the risk predicted by WHO ISH cholesterol method represents the antihypertensive factor in categorizing into high CVD risk category.The alcohol consumption has been proven to be highly associated with the risk of future CVD events (Rehm and Roerecke, 2017). In the present study, alcohol consumption was not a significant predictor in any of the WHO ISH CVD Risk Methods. Alcohol consumption was found to be a significant predictor of WHO ISH CVD Risk (Non-Cholesterol Method) in the multivariate analysis in Indian Kurnool area study and bivariate analysis in Lucknow study (Bandela et al., 2016;Deori et al., 2020). In the Kurnool study method of defining on alcohol consumption was not given and also there had been a high prevelance of alcohol consumption among females in contrast to the present study. In a study from Lucknow India, there was a non-significant association between alcohol consumption with CVD risk predicted by WHO ISH Cholesterol Method. Alcohol consumption was defined as alcohol drinking in the past 30 days in the current study similar to Lucknow study. Multivariate analysis had not been conducted by the Lucknow study investigators. These differences might have given rise to the differences in findings. In the study. In the present study, alcohol consumption was not a significant predictor for CVD risk predicted by any of the Framingham CVD Risk Scores. Similarly, a rural Karnataka study (N = 60), Japanese Minamisomata study (N = 7855) and Baltimore study (N = 1722) found that high Framingham CVD Risk Scores (Original Framingham Equation) were not associated with high alcohol consumption (Allen et al., 2014; Toda et al., 2017 ;George et al., 2017). Only the Japanese study has provided details on alcohol consumption definition which is defined by the frequency of use. However, original or recalibrated Framingham CVD Risk Scores seems to underrepresent risk posed by alcohol consumption status of an individual on CVD risk in the current study sample.

In the present study, only recalibrated Framingham CVD Risk Scores were significantly associated with inadequate vegetable consumption in the Logistic Regression. The Original Framingham CVD scores were associated with inadequate fruit intake significantly. The rural Karnataka study (N = 60) found that the high Framingham CVD Risk Scores (Original Framingham Equation) was not associated with unhealthy dietary intake (adding consumption of junk food) (George et al., 2017). In Kranataka study, no multivariate analysis was conducted. However, in Teheran Lipid and Glucose survey healthy dietary patterns were associated with low risk in multivariate analysis (Ebrahimof et al., 2018). In the Teheran study, Original Framingham CVD Risk Scores were used. This might reflect the risk posed by unhealthy food patterns being represented well by CVD Risk Scores predicted by the Original Framingham Model in Teheran and recalibrated version in the present study.

In the present study, CVD Risk Prediction by only WHO ISH Cholesterol Method was significantly associated with fruit consumption. In a Malaysian study, similar results were obtained for healthy diet in a 24-hour recall (Norhayati et al., 2013). This suggests that the Framingham CVD Risk Scores and WHO ISH Cholesterol Method might represent the risk posed by the inadequate fruit and vegetable consumption. Inadequate fruit and vegetable consumption has been found to be a major risk factor for CVD (Wang et al., 2014). The results have to be interpreted carefully as a 24-hour dietary recall was the only method used to find the adequacy of vegetable and fruit consumption in the present study. In the present study, neither the Framingham Methods nor WHO ISH Methods were associated with physical activity level in the multivariate analysis. However, physical activity level is a known risk factor for CVDs (Wannamethee and Shaper, 2001). Inability of CVD risk prediction methods to represent the physical activity level or sedentary activity duration might imply deficiencies in the risk scores. This may have arisen due to the high physical activity prevalence among the majority which was reflected in the levels of physical activity which are more or less similar in the different risk categories as well. Similarly, A rural Karnataka study (N = 60) found that the high Framingham CVD Risk Scores (original Framingham equation) (> 40%) was not associated with high physical inactivity (using duration of physical activity) (George et al., 2017). Similarly, the Framingham CHD Risk was not significantly associated with physical activity levels in a USA study but was significantly associated with CHD in a study from Finland (LaMonte et al., 2001, Hu et al., 2007). These differences and similarities can arise from the different physical activity levels among different populations. Furthermore, in Finland study physical activity was defined using the frequency of the physical activity conducted. In a rural Panjabi study, high proportion of high CVD Risk predicted by WHO ISH Non-Cholesterol Method was seen among high percentage of physically inactive people where significance of this association was not calculated (Vikramaditya et al., 2017). In a Malaysian study, the physical activity level was not significantly associated with CVD Risk predicted by WHO ISH Cholesterol Method (Norhayati et al., 2013). Similar methodologies followed in these two studies might have resulted in similar findings. The Malaysian study used 7-day physical activity recalls similar to this study. This seems that most of the CVD scores were not sensitive to the risk posed by the lack of physical activity. In the present, waist circumference was found to be significantly associated with the recalibrated Framingham CVD Risk Scores in the multivariate analysis. Similar results were obtained in various studies ( Park and Kim, 2012;Goh et al., 2014). Waist circumference is associated with high CVD Rates (De Koning et al., 2007). This implies the importance of the Framingham Equation after recalibration in representing waist circumference related CVD risk. In the present study CVD risk predicted by WHO ISH CVD Risk Methods were not significantly associated with waist circumference (Table 25, 26). The reason might be the poor performance of WHO ISH Methods in representing waist circumference levels in the current study sample. In a rural Indian study, waist circumference was found to be a significant factor determining CVD risk predicted by WHO ISH CVD Charts in contrast to the present study. Whereas, in the study mentioned above, the direction of the association was not found (no odds ratios were calculated) (Balaji et al., 2018)In the present study, CVD risk predicted by Framingham Original and Recalibrated versions were significantly associated with Triglyceride levels. Similarly, in a Saudi Arabian Military personal study it was found that Triglyceride levels were significantly associated with Framingham CVD Risk Score in the multivariate analysis (Al-Dahi et al., 2013). In a Korean study, bivariate analysis showed that Triglyceride levels were significantly correlated with Framingham CVD Risk Score when using the original Framingham version (Park and Kim, 2012). This shows the representation of risk posed by triglycerides on CVD disease risk through the, the Framingham CVD Risk Score in any population either military or non-military. In the present study, WHO ISH CVD cholesterol method also seems to perform in similar manner. High Triglyceride levels itself is an independent risk factor for CVD risk (Kannel and Vasan, 2009). Therefore, the representation of the risk posed by Triglyceride level might imply high performance of Framingham CVD Risk Methods and WHO ISH cholesterol method regarding indirectly showing CVD risk imposed by high Triglycerides.For further study to compare the performances of each CVD risk score a cohort study is essential.

There were differences in the 10-year risk distribution results by different CVD risk prediction methods. Notable distinctions between the original and recalibrated Framingham scores with were observed. Variable degrees of agreement among the CVD risk predictors were observed and there were significant associations between the non-calculated risk factors with the predicted cardiovascular disease risk.

Recalibration should be conducted using beta coefficient of a cohort study of the study population to provide high accuracy. Performance of CVD Risk calculation methods (calibration and discrimination) should be established from a future cohort study. Associated factors should be assessed with valid tools such as food diaries or food frequency and questionnaires and standard tools measuring alcohol and smoking consumption to improve the validity of data.

BMI

Body Mass Index

CHD

Chronic Heart Disease

CVDs

Cardiovascular diseases

FRS

Org-Original Framingham Risk score

FRS

recal-Recalibrated Framingham Risk score

ISH

International Society of Hypertension

SCORE

Systematic Coronary Risk Evaluation

STEPS

The WHO STEPwise approach to surveillance

WHO

World health Organization

WHO

chole-World Health organization cholesterol

Ethics approval and consent to participate

Ethical approval was obtained from the Ethical Review Committee, Faculty of Medicine, University of Peradeniya. Informed written consent was obtained from all the participants, preserving their freedom to withdraw at any time without any repercussions.

Consent for publication

All the authors and the participants have provided consent for publication.

Competing interests

We declare that there are no competing interests.

Funding

No funding was received.

Author Contribution

SUR and BSR wrote the main text, LIR prepared tables, SUT supervised and reviewed the manuscript

Acknowledgements

No acknowledgements to declare.

Data Availability

I have data available with me in a hard drive and can be provided on request.

Aktas MK, Ozduran V, Pothier CE, Lang R, Lauer MS. Global risk scores and exercise testing for predicting all-cause mortality in a preventive medicine program. JAMA. 2004;292(12):1462–8.
Al-Dahi S, Al-Khashan H, Al Madeer MA, Al-Saif K, Al-Amri MD, Al-Ghamdi O, Al-Helali N, Selim M, Mishriky AM. Assessment of Framingham cardiovascular disease risk among militaries in the Kingdom of Saudi Arabia. Mil Med. 2013;178(3):299–305.
Allen AJ, McNeely JM, Waldstein SR, Evans MK, Zonderman AB. Subjective socioeconomic status predicts Framingham cardiovascular disease risk for whites, not blacks. Ethn Dis. 2014;24(2):150.
Assmann G, Cullen P, Schulte H. Simple scoring scheme for calculating the risk of acute coronary events based on the 10-year follow-up of the prospective cardiovascular Munster (PROCAM) study. Circulation. 2002;105(3):310–5.
Balaji BV, Rajanandh MG, Udayakumar N, Seenivasan P. Prediction of cardiovascular risk in a rural Indian population using WHO/ISH risk prediction charts: a community-based cross-sectional study. Drugs Therapy Perspect. 2018;34:386–91.
Bandela PV, Ambekar NN, Prasad D. Assessment of cardiovascular risk of rural population in Kurnool district using WHO/ISH multivariable risk prediction algorithm.
Cooney MT, Dudina A, D'Agostino R, Graham IM. Cardiovascular risk-estimation systems in primary prevention: do they differ? Do they make a difference? Can we see the future? Circulation. 2010;122(3):300–10.
Cooney MT, Dudina AL, Graham IM. Value and limitations of existing scores for the assessment of cardiovascular risk: a review for clinicians. J Am Coll Cardiol. 2009;54(14):1209–27.
De Koning L, Merchant AT, Pogue J, Anand SS. Waist circumference and waist-to-hip ratio as predictors of cardiovascular events: meta-regression analysis of prospective studies. Eur Heart J. 2007;28(7):850–6.
Deori TJ, Agarwal M, Masood J, Sharma S, Ansari A. Estimation of cardiovascular risk in a rural population of Lucknow district using WHO/ISH risk prediction charts. J Family Med Prim Care. 2020;9(9):4853–60.
Dhingra R, Vasan RS. Age as a risk factor. Med Clin. 2012;96(1):87–91.
Gaziano TA, Pandya A, Steyn K, Levitt N, Mollentze W, Joubert G, Walsh CM, Motala AA, Kruger A, Schutte AE, Naidoo DP. Comparative assessment of absolute cardiovascular disease risk characterization from non-laboratory-based risk assessment in South African populations. BMC Med. 2013;11:1–1.
George CE, Gift N, Mukherjee D, Maddipati T. Law enforcement and cardiovascular risk: Findings of a cross-sectional study from Rural Karnataka. Asian J Med Sci. 2017;8(3):49–54.
Goh LG, Dhaliwal SS, Welborn TA, Lee AH, Della PR. Anthropometric measurements of general and central obesity and the prediction of cardiovascular disease risk in women: a cross-sectional study. BMJ open. 2014;4(2):e004138.
Jones CA, Ross L, Surani N, Dharamshi N, Karmali K. Framingham ten-year general cardiovascular disease risk: agreement between BMI-based and cholesterol-based estimates in a South Asian convenience sample. PLoS ONE. 2015;10(3):e0119183.
WB K. Triglycerides as vascular risk factors: New epidemiologic insights for current opinion in cardiology. Curr Opin Cardiol. 2009;24:345–50.
Khanal MK, Ahmed MM, Moniruzzaman M, Banik PC, Dhungana RR, Bhandari P, Devkota S, Shayami A. Total cardiovascular risk for next 10 years among rural population of Nepal using WHO/ISH risk prediction chart. BMC Res Notes. 2017;10:1–7.
LaMonte MJ, Durstine JL, Addy CL, Irwin ML, Ainsworth BE. Physical activity, physical fitness, and Framingham 10-year risk score: the cross-cultural activity participation study. J Cardiopulm Rehabil Prev. 2001;21(2):63–70.
Leung YY, Lin SC, Lee RS, Lam TH, Schooling CM. Framingham risk score for predicting cardiovascular disease in older adults in Hong Kong. Hong Kong Med J. 2018.
Lemeshow S. Sample size determination in health studies: a practical manual. World Health Organization; 1991.
Mettananda KC, Gunasekara N, Thampoe R, Madurangi S, Pathmeswaran A. Place of cardiovascular risk prediction models in South Asians; agreement between Framingham risk score and WHO/ISH risk charts. Int J Clin Pract. 2021;75(7):e14190.
MINISTRY OF HEALTH 2019. ANNUAL HEALTH BULLETIN 2019.
Nakhaie MR, Koor BE, Salehi SO, Karimpour F. Prediction of cardiovascular disease risk using framingham risk score among office workers, Iran, 2017. Saudi J kidney Dis transplantation. 2018;29(3):608–14.
Norhayati MN, Amirah SS, Husniati YL. Cardiovascular risk: associated factors, assessment and agreement between WHO/ISH risk prediction chart and Framingham Scoring System among Primary Care Patients in Kelantan, Malaysia. Int J Collaborative Res Intern Med Public Health. 2013;5(12):652.
Otgontuya D, Oum S, Buckley BS, Bonita R. Assessment of total cardiovascular risk using WHO/ISH risk prediction charts in three low and middle income countries in Asia. BMC Public Health. 2013;13:1–2.
Park YS, Kim JS. Obesity phenotype and coronary heart disease risk as estimated by the Framingham risk score. J Korean Med Sci. 2012;27(3):243.
Ranasinghe SU, Ekanayake EM, Ranasinghe LI, Tennakoon SU. Recalibration of Framingham risk for a local population of Sri Lanka. BMC Public Health. 2024;24(1):165.
Ranawaka UK, Wijekoon CN, Pathmeswaran A, Kasturiratne A, Gunasekera D, Chackrewarthy S, Kato N, Wickramasinghe AR. Risk estimates of cardiovascular diseases in a Sri Lankan community.
Rehm J, Roerecke M. Cardiovascular effects of alcohol consumption. Trends Cardiovasc Med. 2017;27(8):534–8.
Scheltens T, Verschuren WM, Boshuizen HC, Hoes AW, Zuithoff NP, Bots ML, Grobbee DE. Estimation of cardiovascular risk: a comparison between the Framingham and the SCORE model in people under 60 years of age. Eur J Prev Cardiol. 2008;15(5):562–6.
Schunkert H, König IR, Kathiresan S, Reilly MP, Assimes TL, Holm H, Preuss M, Stewart AF, Barbalic M, Gieger C, Absher D. Large-scale association analysis identifies 13 new susceptibility loci for coronary artery disease. Nat Genet. 2011;43(4):333–8.
Selvarajah S, Kaur G, Haniff J, Cheong KC, Hiong TG, van der Graaf Y, Bots ML. Comparison of the Framingham Risk Score, SCORE and WHO/ISH cardiovascular risk prediction models in an Asian population. Int J Cardiol. 2014;176(1):211–8.
Senaviratna NA. Cardiovascular disease burden in a country: In the context of Sri Lanka. Eur J Med Health Res. 2023;1(3):55–8.
Sivapalaratnam S, Boekholdt SM, Trip MD, Sandhu MS, Luben R, Kastelein JJ, Wareham NJ, Khaw KT. Family history of premature coronary heart disease and risk prediction in the EPIC-Norfolk prospective population study. Heart. 2010;96(24):1985–9.
Tennakoon SU, Kumar BN, Nugegoda DB, Meyer HE. Comparison of cardiovascular risk factors between Sri Lankans living in Kandy and Oslo. BMC Public Health. 2010;10:1–7.
Thulani UB, Mettananda KC, Warnakulasuriya DT, Peiris TS, Kasturiratne KT, Ranawaka UK, Chakrewarthy S, Dassanayake AS, Kurukulasooriya SA, Niriella MA, de Silva ST. (WHO/ISH) cardiovascular risk predictions in Sri Lankans based on findings from a prospective cohort study. PLoS ONE. 2021;16(6):e0252267. Validation of the World Health Organization/International Society of Hypertension.
Toda H, Nomura S, Gilmour S, Tsubokura M, Oikawa T, Lee K, Kiyabu GY, Shibuya K. Assessment of medium-term cardiovascular disease risk after Japan’s 2011 Fukushima Daiichi nuclear accident: a retrospective analysis. BMJ open. 2017;7(12):e018502.
Tulloch-Reid MK, Younger NO, Ferguson TS, Francis DK, Abdulkadri AO, Gordon-Strachan GM, McFarlane SR, Cunningham-Myrie CA, Wilks RJ, Anderson SG. Excess cardiovascular risk burden in Jamaican women does not influence predicted 10-year CVD risk profiles of Jamaica adults: an analysis of the 2007/08 Jamaica Health and Lifestyle Survey. PLoS ONE. 2013;8(6):e66625.
Vikramaditya B, Satija M, Chaudhary A, Sharma S, Girdhar S, Bansal P. A community based cross-sectional study to estimate total cardiovascular risk in rural Punjab. Int J Community Med Public Health. 2017;4(4):1295.
Wang X, Ouyang Y, Liu J, Zhu M, Zhao G, Bao W, Hu FB. Fruit and vegetable consumption and mortality from all causes, cardiovascular disease, and cancer: systematic review and dose-response meta-analysis of prospective cohort studies. BMJ. 2014; 29;349.
Wannamethee SG, Shaper AG. Physical activity in the prevention of cardiovascular disease: an epidemiological perspective. Sports Med. 2001;31:101–14.
World Health Organization. Prevention of Cardiovascular Disease. pocket guidelines for assessment and management of cardiovascular risk. africa: Who/Ish Cardiovascular Risk Prediction Charts for the African Region. World Health Organization; 2007.
WHO. 2020a. Technical package for cardiovascular disease management in primary health care.https://apps.who.int/iris/bitstream/handle/10665/333221/9789240001367-eng.pdf. Accessed 2022.
WHO. 2020b. The top 10 causes of death. https://www.who.int/news-room/fact-sheets/detail/the-top-10-causes-of-death. Accessed 2024.
WHO. 2021. Cardiovascular diseases (CVDs). https://www.who.int/news-room/fact-sheets/detail/cardiovascular-diseases-(cvds)#:~:text=Cardiovascular%20diseases%20(CVDs)%20are%20the,%2D%20and%20middle%2Dincome%20countries. Accessed 2024.
Woodward M, Brindle P, Tunstall-Pedoe H. Adding social deprivation and family history to cardiovascular risk assessment: the ASSIGN score from the Scottish Heart Health Extended Cohort (SHHEC). Heart. 2007;93(2):172–6.

No competing interests reported.

Comparative Analysis of Cardiovascular Disease risk in a local population of Sri Lankans using risk predictors: WHO ISH charts, Original and recalibrated Framingham Equation

Status:

Version 1

Abstract

Background

Methods

Results