**3.3.2 Results**

292 Dyslipidemia - From Prevention to Treatment

**3.3 Study 2 – Usefulness of the lipidic ratios in a low-to-moderate cardiovascular risk population: A sub-analysis of the EDIVA (Estudo de Distensibilidade Vascular) project**  The EDIVA project was an epidemiological study assessing cardiovascular risk through sequential Pulse Wave Velocity measurement (Maldonado, Pereira, Polónia *et al*, 2011), but since serum lipids were available for all the included patients, we re-analyzed the EDIVA database aiming to address the delineated objective: to ascertain the usefulness of The LDL-Cholesterol/HDL-Cholesterol, Total Cholesterol/HDL-Cholesterol and Triglycerides/HDL-Cholesterol ratios in the general population. The study population consisted of 2200 Portuguese nationals (1290 men and 910 women), aged between 18 and 91 years (mean 46.33±13.76 years). Of these, 668 had low cardiovascular risk, and 1532 were patients with hypertension, diabetes and/or dyslipidemia. Individuals defined as having low cardiovascular risk were those who had had no chronic disease, had never been prescribed chronic pharmacological therapy, and had a normal physical exam, electrocardiogram, blood and urine tests, these characteristics having remained unchanged for at least two annual assessments. The patient group was under pharmacological therapy for at least one

The study's aims were explained to all participants and their informed consent was obtained. The methodology used to collect the data was approved by the Portuguese Data Protection Commission and the study was approved by the Ethics Committees of the

This was a prospective, multicenter, observational study monitoring the occurrence of major adverse cardiovascular events (MACE) – death, stroke, transient ischemic attack, myocardial infarction, unstable angina, peripheral arterial disease, revascularization or renal failure. Follow-up of the patients consisted of annual assessments including, blood pressure (BP) measurement, laboratory tests, including serum lipids, and clinical observation. Total cholesterol, triglycerides and HDL cholesterol were measured. LDL cholesterol was obtained by Friedewald's formula (if triglycerides <3.39 mmol/l) or by ultracentrifugation. The LDL-Cholesterol/HDL-Cholesterol, Total Cholesterol/HDL-Cholesterol and Triglycerides/HDL-Cholesterol ratios were calculated in all patients. At each consultation, the subjects' weight and height were measured and body mass index (BMI) was calculated in kg/m2. Blood pressure and heart rate were measured in standard conditions, in a supine position and after a 10-minute resting period, by an experienced operator and using a clinically validated (class A) sphygmomanometer (Colson MAM BP 3AA1-2®; Colson, Paris) (Pereira & Maldonado, 2005).Three measurements were taken and the arithmetic mean was used in the analysis. All participants underwent routine fasting laboratory tests. At the first consultation they filled out a questionnaire concerning relevant personal and

Data from the sample subjects were processed using STATA for Windows, version 11.1. The distribution of the variables was tested for normality using the Kolmogorov-Smirnov test, and for homogeneity of variance by Levene's test. Simple descriptive statistics were used to characterize the sample and the distribution of variables. Cox proportional hazards analysis was used to determine the influence of the lipidic parameters on the occurrence of the specified cardiovascular events. C-Statistics was calculated to address the reliability of the

of the above pathologies.

hospitals involved. Mean follow-up was 2 years.

lipidic parameters as prognostic variables.

family history, smoking habits, alcohol consumption and medication.

**3.3.1 Methods** 

The general characteristics of the study population are summarized in Table 6. Mean age was 46.33±13.77, indicating a relatively young sample, with similar proportions of men and women (59% and 41%, respectively). With regard to cardiovascular risk factors, 52% of the patients were hypertensive, 33% had dyslipidemia and 11% were diabetic; 17% were smokers and 15% had a family history of cardiovascular events. About 37% were medicated for cardiovascular pathologies, with 13.6% of the patients undertaking statins. This factor was controlled in all the multivariable analysis. Mean follow-up is currently 21.42±10.76 months. A total of 50 non-fatal MACE (2.2% of the sample) were recorded, including 27 cases of stroke, 19 of coronary events, 2 of renal failure and 2 of occlusive peripheral arterial disease.


MACE – major acute cardiovascular events; SBP – systolic blood pressure; DBP – diastolic blood pressure; PP – pulse pressure; MAP – mean blood pressure

Table 6. General characteristics of the study cohort, depending on the presence of MACE and conventional cardiovascular risk factors.

Dyslipidemia and Cardiovascular Risk:

*\* p-value<0.01.* 

Lipid Ratios as Risk Factors for Cardiovascular Disease 295

Fig. 3. Adjusted Hazard Ratios for major acute cardiovascular events distributed according to quintiles of the lipid ratios. A) Hazard ratios for quintiles of the LDL-Cholesterol/HDL-Cholesterol ratio; B) Hazard ratios for quintiles of the Triglycerides/HDL-Cholesterol ratio; C) Hazard ratios for quintiles of the Total-Cholesterol/HDL-Cholesterol ratio. The hazard

ratios are multi-adjusted to conventional Framingham cardiovascular risk factors.

Regarding the lipidic profile, patients with MACE presented higher levels of the different lipidic parameters, as illustrated in table 7, in particular the lipidic ratios were significantly higher in patients with MACE (5.76±1.74 versus 6.75±1.98 for Total Cholesterol/HDL-Cholesterol ratio, 3.24±1.32 versus 4.51±1.49 for LDL-Cholesterol/HDL-Cholesterol ratio, 3.17±1.34 versus 4.35±1.67 for Triglycerides/HDL-Cholesterol ratio, *p-value*<0.01). So, overall, the patients with MACE were characterized by an unfavorable metabolic profile compared to the asymptomatic patients.


MACE – major acute cardiovascular events; SBP – systolic blood pressure; DBP – diastolic blood pressure; PP – pulse pressure; MAP – mean blood pressure

Table 7. Lipid profile of the study cohort, stratified for the presence or absence of of MACE.

In the multivariable model analysis, adjusting for all conventional Framingham cardiovascular risk factors (age, sex, diabetes, blood pressure, smoking status and body mass index), the lipids ratios were associated with MACE, with stronger associations than the ones observed for the individual lipidic variables. Overall, the Total-Cholesterol/HDL-Cholesterol was found to be the best single predictor of MACE. In figure 3 we plot the hazard ratios for quintiles of the lipid ratios. A linear increase of the hazard ratios across quintiles of the Total-Cholesterol/HDL-Cholesterol is clearly depicted, while for the other ratios only the upper-extreme quintiles showed an important association with cardiovascular events.

Comparative data of risk association for those in the extreme quintiles of each lipidic variable is presented in figure 4. Of note, one can see that the combination of two individual lipidic components into a single variable provides stronger association with cardiovascular risk, as expressed by the depicted hazard ratios for the lipid ratios. On the other hand, the lipid ratio with the strongest association was the Total-Cholesterol/HDL-Cholesterol ratio, in line with the data depicted in figure 3.

The ROC curve analysis provided the Areas-Under-the-Curve (AUC, equivalent to the Cstatistics) for the different lipid parameter considered in the analysis. The parameters with the biggest AUC were the Total-cholesterol/HDL-cholesterol ratio (AUC=0.703, IC:0.65- 0.77) and the LDL-cholesterol/HDL-cholesterol (AUC=0.701, IC:0.64-0.79).

Regarding the lipidic profile, patients with MACE presented higher levels of the different lipidic parameters, as illustrated in table 7, in particular the lipidic ratios were significantly higher in patients with MACE (5.76±1.74 versus 6.75±1.98 for Total Cholesterol/HDL-Cholesterol ratio, 3.24±1.32 versus 4.51±1.49 for LDL-Cholesterol/HDL-Cholesterol ratio, 3.17±1.34 versus 4.35±1.67 for Triglycerides/HDL-Cholesterol ratio, *p-value*<0.01). So, overall, the patients with MACE were characterized by an unfavorable metabolic profile

**Plasma Total Cholesterol, mg/dl** 221.37±34.01 238.43±36.12 <0,01 **Plasma LDL-Cholesterol, mg/dl** 141.37±31.22 163.26±41.12 <0,01 **Plasma HDL-Cholesterol, mg/dl** 41.22±11.07 36.19±7.28 <0,01 **Plasma Triglicerides, mg/dl** 156.37±34.01 181.43±36.12 <0,01 **LDL-Colesterol/HDL-Colesterol Ratio** 2.98±2.32 4.51±1.49 <0,01 **Total Cholesterol/HDL-Colesterol Ratio** 4.76±2.11 6.75±1.98 <0,01 **Triglicerídeos/HDL-Colesterol Ratio** 3.17±2.32 4.35±1.67 <0,01

MACE – major acute cardiovascular events; SBP – systolic blood pressure; DBP – diastolic blood

Table 7. Lipid profile of the study cohort, stratified for the presence or absence of of MACE. In the multivariable model analysis, adjusting for all conventional Framingham cardiovascular risk factors (age, sex, diabetes, blood pressure, smoking status and body mass index), the lipids ratios were associated with MACE, with stronger associations than the ones observed for the individual lipidic variables. Overall, the Total-Cholesterol/HDL-Cholesterol was found to be the best single predictor of MACE. In figure 3 we plot the hazard ratios for quintiles of the lipid ratios. A linear increase of the hazard ratios across quintiles of the Total-Cholesterol/HDL-Cholesterol is clearly depicted, while for the other ratios only the upper-extreme quintiles showed an important association with

Comparative data of risk association for those in the extreme quintiles of each lipidic variable is presented in figure 4. Of note, one can see that the combination of two individual lipidic components into a single variable provides stronger association with cardiovascular risk, as expressed by the depicted hazard ratios for the lipid ratios. On the other hand, the lipid ratio with the strongest association was the Total-Cholesterol/HDL-Cholesterol ratio,

The ROC curve analysis provided the Areas-Under-the-Curve (AUC, equivalent to the Cstatistics) for the different lipid parameter considered in the analysis. The parameters with the biggest AUC were the Total-cholesterol/HDL-cholesterol ratio (AUC=0.703, IC:0.65-

0.77) and the LDL-cholesterol/HDL-cholesterol (AUC=0.701, IC:0.64-0.79).

**No MACE**  (n=2150)

**MACE**  (n=50)

*p-value*

compared to the asymptomatic patients.

pressure; PP – pulse pressure; MAP – mean blood pressure

cardiovascular events.

in line with the data depicted in figure 3.

*\* p-value<0.01.* 

Fig. 3. Adjusted Hazard Ratios for major acute cardiovascular events distributed according to quintiles of the lipid ratios. A) Hazard ratios for quintiles of the LDL-Cholesterol/HDL-Cholesterol ratio; B) Hazard ratios for quintiles of the Triglycerides/HDL-Cholesterol ratio; C) Hazard ratios for quintiles of the Total-Cholesterol/HDL-Cholesterol ratio. The hazard ratios are multi-adjusted to conventional Framingham cardiovascular risk factors.

Dyslipidemia and Cardiovascular Risk:

various protagonists over the clinical decision frame.

Lipid Ratios as Risk Factors for Cardiovascular Disease 297

conventional lipid components (Walldius & Junger, 2006). In contrast, the concentration of apo A-1 translates the number of anti-atherogenic particles contained in the HDLcholesterol, thus enclosing the conceptual framework of apoB/apoA-1 ratio as a measure of the ratio of atherogenic particles versus anti-atherogenic particles transported in the blood. Despite the growing enthusiasm about the potential of these emerging parameters for their best performance in the definition of cardiovascular risk, there still remain some questions that limit their dissemination in clinical practice. The central question is very practical, and focuses on the cost-benefit relation associated with a change in the traditional clinical approach. In fact, it is not yet clear whether the superiority of these new lipid parameters over the more conventional ones for risk stratification is enough to justify the additional cost inherent to their laboratory determination (Pischon, Girman, Sacks *et al*, 2005). Furthermore, despite the current literature supporting apolipoproteins as better predictors of cardiovascular events, its use may not be the most practical operational perspective. Moreover, it is not yet clear whether the replacement of conventional parameters for emerging ones will translate into clear clinical benefit, or if, conversely, it will confuse the

In contrast to this line of argument, several studies have also emerged affirming quite clearly the advantages of using lipid ratios, based on conventional parameters, such as those studied in this work. This is based on the fact that, on the one hand, they add cardiovascular risk discriminative capacity to the individual lipid parameters, and on the other, they are more favorable than the apolipoproteins considering cost and immediate operationalization (Gotto, Whitney & Stein, 2000). As mentioned earlier, several studies have shown fairly consistently that changes in these ratios are favorable indicators of cardiovascular disease risk, above the absolute levels of individual lipids (Natarajan, Glick, Criqui *et al*, 2003; Kannel, 2005). Accumulating evidence in this regard is quite broad, spreading over several clinical frameworks (Manninen, Tenkanen, Koskinen *et al*, 1992; Kannel, 2005; Cullen, Schulte, Assmann *et al*, 1997; Stampfer, Sacks, Salvini *et al*, 1991; Gaziano, Hennekens, O'Donnell *et al*, 1997; Packard, Ford, Robertson *et al*, 2005). The results presented here clearly fall into this line, reinforcing the belief in the superiority of the lipid ratios, particularly the Total-Cholesterol/HDL-Cholesterol and the LDL-Cholesterol/HDL-Cholesterol ratios, over the classic lipid parameters, predicting peripheral arterial disease in hypertensive patients (in a high cardiovascular risk) and predicting future major cardiovascular events (including stroke and myocardial infarction) in a low-to-intermediate cardiovascular risk population. One of the curious aspects extracted from the second presented study was the existence of a linear relationship for the Total-Cholesterol/HDL-Cholesterol ratio with the risk of MACE, something not apparent in the LDL-Cholesterol/HDL-Cholesterol ratio. This same result was reproduced in the Quebec Cardiovascular Study, in which more than 2.000 middleaged men were followed for 5 years, monitoring the occurrence of major cardiovascular events (Lemieux, Lamarche, Couillard *et al*, 2001). The lipid parameters with better performance in predicting risk in this study were the Total-Cholesterol/HDL-Cholesterol ratio and the LDL-Cholesterol/HDL-Cholesterol ratio, although only the first stated ratio expressed a linear relationship with risk. One possible explenation for this result is metabolic in nature. In fact, it is well documented that patients with dyslipidemia showing high triglycerides and low HDL-cholesterol (generally patients with abdominal obesity and insulin resistance), often have marginal or even normal levels of LDL-Cholesterol (Lamarche, Després, Moorjani *et al*, 1996). Moreover, LDL-Cholesterol concentrations are often estimated indirectly from 3 measurements (Total-Cholesterol, Triglycerides and HDL-

Fig. 4. Adjusted Hazard Ratios for major acute cardiovascular events amongst those in the extreme quintiles of each considered lipidic parameter. The hazard ratios are multi-adjusted to conventional Framingham cardiovascular risk factors.
