Results
169 patients without previous CVDs; 130(81.1%) females, mean 65 ±6.9 years (range 47-78 years) were studied. Baseline characteristics, risk factors, and medication use of the sample are shown in Table 1. Of the total population, 56.2% were 65 years or older. The majority, 98.8% were Singhalese. Risk factor distribution among the study participants was, hypertension 66.9%, hyperlipidemia 89.9%, diabetes mellitus 46.7%, smoking 2.4%, and obesity 8.3%. Patients were on medications; antihypertensives 65.7%, lipid-lowering medications 89.9%, anti-diabetic medications 46.2%, and antiplatelet medications 15.5% among the total population. Men in comparison to women were older ( 68 ± 4.79 and 64 ± 7.03 years, p <0.0001 respectively), smoked more (12.5%, 0.0%, p <0.0001) and were less likely to be on lipid-lowering medications (77.4%, 92.7%, p=0.019).
Table 1 Baseline characteristics of the study population
Comparison of risk factors used in the calculation of Framingham and WHO/ISH scores and mean FRS of men and women are shown in Table 2. There was no significant difference in the history of diabetes mellitus, use of anti-hypertensive medications, and measured risk factors like BMI, SBP, TC and HDL levels between men and women. The two groups were only different from age and smoking status. However, the mean FRS of men were significantly higher than that of females with both BMI-based (male 28.94 ± 3.17, female 17.10 ±8,62) and cholesterol-based (male 26.47 ± 4.99, female 13.86 ± 8.25) models.
Table 2 CV-risk factors used in risk calculations and mean Framingham risk scores by sex
Patients were categorised into low(<20% ) and high(\(\geq\)20%) CV-risk groups on risk predictions (Table 3). 80(47.3%), 62 (36.7%), 18 (10.7%), 16 (9.5%), of the participants were predicted high risk by FRS BMI-based, FRS cholesterol-based, WHO/ISH without-cholesterol and WHO/ISH with-cholesterol models, respectively. Agreement between different risk models in categorizing patients into low and high-risk groups was studied using Cohen’s kappa statistics (Table 3).
Table 3 10-year CV-Risk stratification of the sample with different risk models and inter-rater agreement
The two versions of FRS models; BMI-based and cholesterol-based were in good agreement in stratifying patients into high and low-risk groups, κ = 0.736, p<0.0001. Similarly, the two versions of WHO/ISH models without-cholesterol and with-cholesterol were also in good agreement in stratifying patients into high and low-risk groups; κ = 0.804, p<0.0001. However, the agreement between, FRS BMI-based model and WHO/ISH without-cholesterol model in stratifying patients into high and low-risk groups was fair; κ = 0.234, p<0.0001 and FRS BMI-based risk estimates were higher than WHO/ISH without-cholesterol estimates. Furthermore, the agreement between, FRS cholesterol-based model and WHO/ISH with-cholesterol model in stratifying patients into high and low-risk groups was also fair; κ = 0.306, p<0.0001 and FRS cholesterol-based risk estimates were higher than WHO/ISH with-cholesterol estimates.