Comparative validation of low-density lipoprotein cholesterol estimation formulas in older Georgian adults

Low-density lipoprotein cholesterol (LDL-C) is a critical marker for cardiovascular risk assessment. Although direct measurement offers high accuracy, it is often cost-prohibitive and impractical for routine use in low- and middle-income countries. Multiple formulas, including those by Friedewald, de Cordova, and Chen, have been proposed to estimate LDL-C, though their accuracy varies across populations. This study evaluated the performance of eight LDL-C estimation formulas against direct measurement in a predominantly older adult Georgian cohort.

We retrospectively analyzed lipid profiles from 500 adults with complete panels, stratified by triglyceride (TG) levels, high-density lipoprotein cholesterol (HDL-C), total cholesterol (TC), and age. LDL-C was estimated using eight formulas and compared with direct LDL-C assays. The study adhered to the Helsinki Declaration and was approved by the Bioethics International Committee of Petre Shotadze Tbilisi Medical Academy.

Substantial variability was observed across formulas. Friedewald and Chen showed minimal underestimation, aligning well with direct measurements, particularly at moderate TG levels. The de Cordova formula maintained stable accuracy across TG strata, including borderline hypertriglyceridaemia. The Ahmadi formula, originally developed for mmol/L, produced significant overestimation in mg/dL units and was excluded from threshold-based analyses. Sensitivity testing using CLIA's dual total allowable error (TEa) thresholds (±12 % or 12 mg/dL) improved agreement for all formulas, especially at low LDL-C levels.

Friedewald and de Cordova offer reliable, cost-effective LDL-C estimates for older adults. Formula selection should account for TG levels, demographics, and analytical context. Broader validation in diverse cohorts is needed to enhance generalizability.