Preanalytical factors for non-HDL cholesterol measurements/ Serum lipid profiles including non-high density lipoprotein cholesterol levels in Turkish school-children/Non-HDL kolesterol olcumlerinde preanalitik faktorler/Turk okul cocuklarinda serum lipid profili ve non-HDL kolesterol duzeyleri.

Dear Editor,

I have read the paper entitled "Serum Lipid Profiles Including

non-High Density Lipoprotein Cholesterol Levels in Turkish

School-Children" by Ugar et al. (1) with great interest. I agree with the

authors but two major points in the paper should be re-evaluated:

(1) The authors stressed that the measurement of non-HDL

cholesterol does not require overnight fasting.

(2) The authors found a strong positive correlation between non-HDL

cholesterol and LDL cholesterol in the study population

In clinical laboratories, preanalytical factors such as, hemolysis,

icterus, drugs, and especially lipemia are important. To prevent lipemia,

patients should be starved prior to drawing blood (2). The authors

suggested that starving is not necessary for measuring non-high density

lipoprotein (HDL) cholesterol, and based their claim on a previous study,

which had obtained non-HDL cholesterol by calculation using total

cholesterol and HDL (3). Their claim is definitely not valid for two

important reasons: preanalytical conditions and the inappropriate

comparison of reference intervals.

Lipemia is an important interferentfor common tests, including total

cholesterol and HDL-cholesterol measurements, and cannot be ignored.

Furthermore, even if we could measure total cholesterol and HDL-cholesterol

or any other biochemical parameters correctly with

sophisticated methods, and solve all analytical problems, we cannot omit

starvation prior to analysis. We compare patients' test results with

reference values and make our decisions on the basis of this comparison.

For a realistic and objective comparison in clinical practice, the

preanalytical condition of patients should be the same as that of the

reference population. To determine the reference interval in the

reference population, we generally take samples at the same time of day

(usually in the morning and preferably between 07:00 and 09:00 AM)

under standard preanalytical conditions. All subjects should abstain from

food after 2200 hr the evening before specimen collection (4). If we

measure non-HDL cholesterol in non-starved patients, we should

compare patient results with a reference interval that was obtained from

a non-starved reference population. Therefore, overnight fasting is

necessary for measuring non-HDL cholesterol as well as for other

biochemical tests for a realistic comparison with reference values, even

if no analytical problems exist.

The second important point is that the authors found a strong

positive correlation between non-HDL cholesterol and low-density

lipoprotein (LDL) cholesterol in their study population. In my opinion, this

is not a special condition for Turkish school-children and these results

could be seen in any population. If we examine the relationship between

lipid parameters, we can see the situation clearly.

If we neglect serum intermediate density lipoprotein (IDL) and

lipoprotein(a) [Lp(a)] concentrations, we can write

non-HDL cholesterol=LDL cholesterol + VLDL cholesterol [1].

We can see from equation 1 that the concentration of non-HDL

cholesterol is dependent on the concentration of LDL cholesterol, i.e., LDL

cholesterol is not an independent variable. Both very low-density

lipoprotein (VLDL) and LDL contain apolipoprotein B100; consequently, we

may assume that the strong positive correlation between LDL and

non-HDL cholesterol levels is to be expected and is not informative. I think

the correlation between non-HDL cholesterol and clinical parameters

would be more informative than the correlation with LDL-cholesterol.

Abdurrahman Coskun

Department of Clinical Biochemistry School of Medicine, Duzce University, Duzce, Turkey

References

(1.) Ucar B, Kilic Z, Dinleyici EC, Colak O, Gunes E. Serum lipid profiles including non-high density lipoprotein cholesterol levels in Turkish school-children. Anadolu Kardiyol Derg 2007; 7: 415-20.

(2.) Coskun A. Total variation of LDL-cholesterol and apolipoprotein B / Serum apolipoprotein B is superior to LDL-cholesterol level in predicting incident coronary disease among Turks. Anadolu Kardiyol Derg 2007; 7: 331-2.

(3.) Srinivasan SR, Myers L, Berenson GS. Distribution and correlates of non-high-density lipoprotein cholesterol in children: The Bogalusa Heart Study. Pediatrics 2002;110: e29.

(4.) Solberg HE. Establishment and use of reference values. In: Burtis CA, Ashwood ER, Bruns DE, editors. Tietz Textbook of Clinical Chemistry and Molecular Diagnostics. St. Louis, MO: Elsevier Saunders; 2006. p. 425-48.

Address for Correspondence/ Yazisma Adresi: Dr. Abdurrahman Coskun Duzce University, School of Medicine, Department of Clinical Biochemistry 81620 Konuralp, Duzce, Turkey

Phone: [phone] E-mail: [e-mail]

Authors Reply

Dear Editor,

We appreciate the author for his/her interest in reference to our article entitled "Serum lipid profiles including non-High Density Lipoprotein Cholesterol Levels in Turkish School-Children" that was recently published in Anadolu Kardiyoloji Dergisi (1). The author drew attention to the potential effect of pre-analytical, analytical, and biological factors on serum lipid and lipoprotein parameters.

While the measurement of low-density lipoprotein cholesterol (LDL-C) is widely considered as the gold standard as part of lipid profile for screening of coronary artery disease, also has some limitations. The LDL-C is typically calculated with the Friedwald formula, which requires the measurement of total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C) and triglycerides (TG) (2). Because triglyceride levels can fluctuate significantly in relation to the patient's fasting status, estimation of LDL-C levels requires overnight fasting. Furthermore, these calculated LDL-C values include also intermediate-density lipoprotein cholesterol (IDL-C) levels and lipoprotein (a) to varying degrees. For these limitations, the measurement of non-high-density lipoprotein cholesterol (non-HDL-C) levels (with total cholesterol minus HDL-C formula) has been proposed as a better screening tool for coronary artery disease risk assessment and the measurement does not require overnight fasting (3-4). Although we have written that the measurement of non-HDL cholesterol do not require overnight fasting, we had obtained venous blood samples from children after at least 12 hours fasting in the morning hours as we have noted in the methods section of our article. So, there is not any limitation for making comparison between our results and those of any other studies including starved subjects. We did not compare lipid levels of our study population with those of other populations and we estimated our own mean and percentile values. On the other hand, we think that the percentile values for all lipid parameters which were reported in our study may be used as reliable reference values especially for Turkish children, because our study includes quite a large population including 2896 children (1467 girls and 1429 boys) that can be representative for Turkish school-children.

The aim of our study was to measure plasma lipids and lipoprotein levels, to evaluate the prevalence of dyslipidemia according to non-HDL-C levels besides the other conventional lipid parameters and also to evaluate serum lipid and lipoprotein levels according to age, gender difference and living areas in school-children. The author suggests that the strong correlation between LDL-C and non-HDL-C levels is to be expected because the concentration of non-HDL-C is dependent on the concentration of LDL-C. In our study serum non-HDL-C levels correlated with serum TC, TG, HDL-C, LDL-C and VLDL-C levels in both sexes. However, we highlighted the strong correlation between serum non-HDL-C and LDL-C levels, because of serum LDL-C level was classical indicator for cardiovascular diseases in children as well as in adults.

The author suggests that the correlation between non-HDL-C levels and clinical parameters would be more informative. In our study, we also evaluated and reported correlations of non-HDL-C levels with age, weight, height, total body fat percentage, body mass index and physical activity status for both sexes. We found positive correlations with age in both sexes, and with height, weight and total body fat percentage in boys but not in girls. Our findings indicate that the unfavorable situations affecting both anthropometric values (especially related with fatness) and lipid parameters, which lead to the male predominance in cardiovascular diseases in adults begin in childhood.

Birsen Ugar, Zubeyir Kilic, Ener Cagri Dinleyici *, Omer Colak **, Erdogan Gunes *

From Department of Pediatric Cardiology, Pediatrics, and Biochemistry", Faculty of Medicine, Eskisehir Osmangazi

University, Eskisehir, Turkey

References

(1.) Ucar B, Kilic Z, Dinleyici EC, Colak O, Gunes E. Serum lipid profiles including non-high density lipoprotein cholesterol levels in Turkish school-children. Anadolu Kardiyol Derg 2007; 7: 415-20.

(2.) Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972; 18: 499-502.

(3.) Srinivasan SR, Myers L, Berenson GS. Distribution and correlates of non-high-density lipoprotein cholesterol in children: the Bogalusa Heart Study. Pediatrics 2002;110: e29.

(4.) Farwell WR, Sesso HD, Buring JE, Gaziano JIM. Non-high-density lipoprotein cholesterol versus low-density lipoprotein cholesterol as a risk factor for a first nonfatal myocardial infarction. Am J Cardiol 2005; 96:1129-34.