Determining a new formula for calculating low-density lipoprotein cholesterol: data mining approach.

Area of Science:

• Clinical Chemistry
• Cardiovascular Disease Risk Assessment

Background:

• Low-density lipoprotein cholesterol (LDL-C) is a key risk factor for coronary heart disease.
• The Friedewald equation for estimating LDL-C has limitations, particularly with triglyceride (TG) levels exceeding 400 mg/dL.

Purpose of the Study:

• To develop and validate a new formula for estimating LDL-C (LDL-Cal).
• To assess the correlation between the new LDL-Cal and directly measured LDL-C (LDL-Direct).
• To extend the applicability of LDL-C estimation to higher triglyceride levels.

Main Methods:

• A dataset of 1786 individuals was used to construct the LDL-Cal formula, dividing TG levels into six groups.
• A pace regression model was employed to generate candidate formulas and determine correlation coefficients (r).
• An external dataset of 666 individuals was used for validation.

Main Results:

• A novel LDL-Cal formula was generated, showing a strong correlation (r=0.9769) with LDL-Direct for TG levels up to 1000 mg/dL.
• Validation using an external dataset confirmed a high correlation (r=0.971) between the new LDL-Cal and LDL-Direct.
• The new formula demonstrated effectiveness beyond the TG limitation of the Friedewald equation.

Conclusions:

• A new, highly accurate formula for LDL-C estimation has been developed.
• This formula overcomes the triglyceride limitations of the Friedewald equation.
• The new LDL-Cal formula has potential for routine use in clinical laboratories.