Evaluation of lipid parameters and bioindices in patients with different stages of chronic renal failure

Background/Aim. Cardiovascular morbidity and mortality are markedly increased in chronic renal failure (CRF). The aim of this study was to evaluate lipid parameters and bioindices in patients with different stages of CRF. Methods. In 46 hemodialysed (HD), 50 CRF patients with II, III and IV stage of CRF (non-HD) and 48 control subjects triglycerides (TG), total cholesterol (C), HDL-C, urea, creatinine, creatinuria (standard biochemical methods), apolipoprotein (apo) AI, apo B, lipoprotein(a), cystatin C (immunoturbidimetric method) were evaluated, and LDL-C, non-HDL-C, LDLC/HDL-C, non-HDL-/HDL-C, TG/HDL-C, and new bioindices, LTI (lipid tetrad index), logLTI, LPI (lipid pentad index), logLPI, AIP (atherogenic index of plasma), and creatinine clearance were calculated. Results. There were significant differences in the levels of TG, HDL-C, LDL-C, nonHDL-C, total C and apo A-I between the HD and non-HD patients, and the HD patients and the controls. LTI and LPI were significantly higher in the HD and non-HD patients compared to the controls (p < 0.05), without a good separation by the Box-Whisker plots. The values of TG/HDL-C ratio and AIP were significantly higher in the HD and nonHD-patients compared to the controls (p < 0.05), and significantly higher in the HD compared to non-HD patients (p < 0.05). AIP > 0.11 was found in 71.7% of the HD, 56% of non-HD and 31.3% of the controls. Conclusion. Among lipid parameters and indices, AIP and TG/HDL-C ratio are most suitable for evaluation of lipid disturbances in different stages of CRF. In addition to, non-HDL-/HDL-C, and apoB/A-I ratios, apo A-I, HDL-C and TG are important markers in HD patients. Non-HDL-C is not a suitable marker. LTI and LPI need to be further investigated.


Introduction
Epidemiological data indicate that approximately 10% of adult population have some form of chronic renal disease that eventually may progress into a complete loss of kidney function 1 .However, before that, the majority of patients will die from fatal cardiovascular events, particularly from coronary heart disese 2 .Moreover, cardiovascular mortality is 10 to 30 times higher in hemodialysis patients compared to general population 3 .
Dyslipidemia is a well-known traditional risk factor for premature atherosclerosis.According to the current guidelines for the diagnosis and treatment of cardiovascular disease (CVD), it is considered a major risk factor 4 .In renal patients, dyslipidemia is being present in 40-60% 5 .
Lipid disturbance in chronic renal failure (CRF) are markedly expressed and specific.Due to the dysregulation of numerous enzymes, apolipoproteins and receptors, maturation and metabolism of HDL and metabolism of triglyceride-rich lipoproteins are decreased 5,6 .A typical lipid profile in CRF is thus characterized by a combination of quantitative and qualitative abnormalities, including hypertriglyceridemia, decreased HDL-C levels, normal, mildly increased, or even mildly decreased total and LDL cholesterol levels, and an abnormal lipid subfraction profile with a predominance of atherogenic low-dense LDL (sd LDL) and HDL particles 7 .The retention of lipoprotein particles of modified structure and size occurs already in the early, pre-dialysis stage, before the elevation of plasma lipids, and persists even after successful transplatation 8 .Plasma lipids and apolipoproteins may hence be unreliable predictors of the cardiovascular risk in CRFinduced dyslipidemia, especially in the early stages of kidney disease.
The mechanism underlying the atherogenic lipoprotein profile associated with renal failure involves delicate metabolic interrelations within the whole lipoprotein system, which is typically present when major traditional risk factors such as hypertension, glucose intolerance and abdominal obesity co-occur 9,10 .
For such patient populations calculations of novel bioindices have recently been proposed, such as lipid tetrad index (LTI), lipid pentad index (LPI) 11,12 and atherogenic index of plasma (AIP) 13 , whose formulas comprise several lipid parameters (total cholesterol, triglycerides, HDL-C, apo A, apo B and Lp(a)).In addition, AIP is an indirect indicator of sdLDL levels 13 .
Considering that there are no data about these approaches in CRF patients, the aim of this research was to evaluate the new lipid bioindices (LPI, LTI and AIP) and traditional lipid parameters and ratios (total cholesterol, triglycerides, LDL-C, HDL-C, non-HDL-C, Lp(a), apo A-I and B, apo B/A-I, LDL-C/HDL-C, non-HDL-/HDL-C, triglycerides/HDL-C) in patients with different stages of CRF (hemodialysed and non-hemodialysed).

Methods
The study was carried out in the Clinical Center of Vojvodina, Novi Sad, Serbia in the period January-December 2009.The study had previously been approved by an institutional ethics committee and all the subjects included had consented to participate.A total of 144 subjects were included -96 with CRF [46 on chronic hemodialysis (HD) and 50 non-hemodialysed (non-HD) patients in different stages of CRF] and 48 controls with normal creatinine clearance (CrCl) and without hypolipemic or antihypertensive therapy.
The inclusion criteria were CRF of various etiologies and stages and chronic hemodialysis treatment lasting more than 18 months.Blood samples were collected on the first day of weekly hemodialysis, before connecting the patient with the dialyser and administering heparin.Patients with nephrotic syndrome, acute infection, liver disease, malignancies, previous or recent myocardial infarction, stroke or peripheral arterial disease were excluded from the study.
Analyses were performed immediately after bloodsampling.Serum levels of creatinine, urea, total C, and HDL-C, TG and creatinuria were measured using the standard biochemical methods, and apo A-I and B, Lp(a) and cystatin C by immunoturbidimetry.
Creatinine clearance was calculated by the formula: CrCl = U Cr x 24 h urine volume (ml) / S Cr x 1440 (min) ; Note: U Cr urine creatinine (μmol/L) and S Cr -serum creatinine (μmol/L) The obtained values were normalized to 1.73 m 2 body surface area, sex and age.A deviation above 10% from the expected value in the past 6 months was taken as a criterion for the existence of CRF.
We calculated the values of LDL-C, non-HDL-C, and traditional indICes LDL-/HDL-C, non-HDL-/HDL-C, TG/HDL-C, and new bioindexes Lipid tetrad index (LTI) = TG x total C x Lp(a)/HDL-C; Lipid pentad index (LPI) = TG x total C x Lp(a) x apo B/A-I; Atherogenic index of plasma (AIP) = log (TG/HDL-C).All the parameters were translated into the SI units.The patients were classified accordingly into three risk categories for AIP: low < 0.11, intermediate 0.11-0.21,and high > 0.21 14 .Since the indices LPI and LTI did not have normal distributions, we used variables logLPI and logLTI.
In addition, body mass index (BMI) was calculated for all subjects 15 .
Descriptive statistics, including median, arithmetic mean, standard deviation (SD) and standard error (SE) were used to describe the studied parameters.The differences in distributions of individual parameters between study groups were analyzed using the parametric Student's t-test, or the non-parametric Mann-Whitney test in case a distribution showed a significant deviation.Linear regression analysis and Pearson coefficient of linear correlation were used to study the correlation between variables.The differences between the groups were illustrated using Box and Whisker plots and empirical distribution functions.Statistical analysis was performed using the Statistica 8.0 software.A value of p < 0.05 was considered statistically significant.

Results
The main characteristics of the study subjects are shown in Table 1.There were significant differences in age and BMI between the HD and non-HD patients.
Laboratory parameters and the calculated indices are presented in Tables 2 and 3.
Total C, TG, HDL-C, LDL-C, non-HDL-C, apo A-I, and non-HDL-/HDL-C levels, and apo B/A-I and TG/HDL-C ratios were significantly different in the HD patients compared with the other two groups.TG/HDL-C ratio was significantly higher in the non-HD patients compared with controls.Distribution of patological values of LDL-/HDL-C, apoB/A-I and non-HDL-/HDL-C ratios in non-HD and HD patients are presented in Table 4.
There was a significant difference in the mean logLPI and logLTI between the controls and the two groups of patients, but not between the HD and non-HD patients (Table 2).This was supported by the non-parametric Mann-Whitney test for LPI and LTI.However, considering great dispersion around both indices their values overlapped.This was illustrated on empirical function distribution plots in Figures 1  and 2. The values below the upper limit of the confidence interval for the mean of the log LTI (6.23) were recorded in 70.83% of the controls, 46% of the non-HD patients and 36.9% of the HD patients.The values below the upper limit of the confidence interval for the mean of the log LPI (9.80) were recorded in 68.75%, of the controls, 46% of the non-HD patients and 39.13% of the HD patients.

Box Plot of
Similar results were obtained for AIP, however, the significance of the difference in the means in the controls and the HD patients (p < 0.001) was higher than in the case of LPI and LTI (p < 0.05).Box-plots (Figure 3) showed that most HD patients had AIP over 0.11 (71.7%), which was significantly higher (p = 0.0001) compared with the control group, where 31.3% of the subjects had this finding (Table 3).
In the non-HD patients, however, the percentage was insignificantly higher in comparison with the control group (46%, p = 0.067).

Discussion
Cardiovascular morbidity and mortality are present already in mild renal impairment 16,17 and increased in endstage renal disease 18 .
Among many factors contributing to accelerated atherosclerosis in CRF, dyslipidemia is found in 40-60% patients 5 .Regardless the etiology, patients with CRF develop complex qualitative and quantitative abnormalities, predominantly hypertriglyceridemia and HDL-decrease.In the present study, the highest triglyceride levels, the lowest HDL-C levels, and the highest values of TG/HDL-C ratio were observed in HD patients.Beside this, the values of TG/HDL-C ratio, that is one of the most potent predictors of cardiovascular disease 8 , were significantly higher in the non-HD patients compared with the controls.More than half of the HD group had high-risk HDL-C levels, which is similar to the results of other authors 19,20 .
Total and LDL cholesterol levels were lowest in the HD patients, compared to the controls and non-HD patients (p < 0.001), whereas the levels in the non-HD patients were similar to those of the controls.Most previous studies have also reported similar or lower values 9,19,[21][22][23] .Unlike general population, lower plasma cholesterol in CRF has been associated with a higher cardiovascular mortality 24 .
Similarly to the results of Schreier at al. 25 , the results of our study show that non-HDL-C may not be appropriate marker for risk assessment among CRF patients.The level of non-HDL-C in the HD patients was significantly lower compared with the controls, with no differences among the other groups.The level of non-HDL-C/HDL-C ratio was significantly higher in the HD patients compared with the others groups, mainly due to low HDL-C levels.
More recent research indicates that apolipoproteins are more effective atherogenic markers than plasma lipids: apoA-I is a useful summary index of the antiatherogenic properties of HDL [26][27][28] , apoB of total atherogenic particle number [29][30][31][32] , and apo B/A-I ratio is better than any cholesterol ratio 26,33,34 .In our study, apoA-I was significantly lower only in the HD patients, whereas apoB did not differ significantly, similarly to the results of Alabakovska et al. 19 .The frequency of pathological ratio apoB/A-I was significantly higher than LDL-/HDL-C in both groups of patients in our study.
Although relations between Lp(a) and renal functional status in patients with CRF have been explored in numerous studies, the results obtained are contradictory 19,35,36 .In our study, Lp(a) levels in the HD and non-HD patients did not differ significantly from the controls.Furthermore, there was no significant correlation between Lp(a) levels and CrCl (r = 0.12, p = 0.51), contrary to some other studies dealing with early stages of CRF 37 .
Literature data on LTI and LPI in CRF patients are scarce.Despite a significant difference in bioindices between healthy controls and patients with CRF, we did not obtain adequate delineations, which prevented us from determining reference values.An explanation could lie in similar distributions of Lp(a) levels and the existence of extreme pathological values of certain lipid parameters in both controls and patients that is characteristic for our geographic area.
Furthermore, there are no literature data on AIP in patients with CRF.We found the values of > 0.11, which indicated moderate and high atherogenic risk, in 31.3% of the control subjects, 56% of non-HD patients and 71.7% of HD patients.It is significantly different between the groups, particularly controls and HD patients.These findings were expected, since CRF is characterized by the predominance of atherogenic sdLDL particles.

Conclusion
In conclusion, TG, HDL-C, apo A-I, non-HDL-/HDL-C, and apoB/A-I ratios, are important lipid markers only in HD patients.Non-HDL-C is not a suitable marker in CRF patients.Calculations of bioindices LPI and LTI did not show significant benefits in our study population.In contrast, significantly increased pathological values of AIP that we registered in our study in both HD and non-HD patients suggest that this index may have a potential application in routine clinical practice as an indirect indicator of sdLDL levels.Besides AIP, TG/HDL-C ratio could also be a suitable marker for evaluation of lipid disturbances in different stages of CRF.