Comparing Low-density Lipoprotein Cholesterol Population Estimates Using Different Predictive Equations: National...

Area of Science:

• Cardiovascular Diagnostics
• Biochemical Analysis
• Epidemiology

Background:

• The Friedewald equation, used since 1972, estimates low-density lipoprotein cholesterol (LDL-C) but has accuracy limitations for triglyceride levels.
• Newer equations, Martin (2013, extended 2021) and Sampson (2020), offer improved accuracy across a wider range of triglyceride levels.
• The choice of LDL-C calculation method can influence clinical management and population health estimates.

Purpose of the Study:

• To compare LDL-C calculations using Friedewald, Martin, and Sampson equations in a U.S. adult population.
• To assess the impact of different LDL-C equations on classification across clinical cut points.
• To evaluate the agreement between these equations in a nationally representative sample.

Main Methods:

• Utilized data from 4,461 adults (2015-2018 National Health and Nutrition Examination Survey).
• Assessed classification agreement into LDL-C categories using kappa statistics across demographic subgroups.
• Conducted a sensitivity analysis for Martin and Sampson equations with triglycerides <800 mg/dL.

Main Results:

• Population-level LDL-C distributions were similar across the three equations for adults aged 20+.
• Prevalence estimates for LDL-C <70 mg/dL ranged from 9.8% to 10.0%.
• Prevalence estimates for LDL-C >190 mg/dL were below 3% for all equations, with very good agreement (kappa >0.8) observed across subgroups.

Conclusions:

• The Friedewald, Martin, and Sampson equations yield comparable U.S. population-level LDL-C distributions.
• The selection of LDL-C calculation equation has a minimal effect on national estimates for adults.
• These findings support the consistency of LDL-C categorization across different calculation methods in large-scale studies.