A New Equation Based on the Standard Lipid Panel for Calculating Small Dense Low-Density Lipoprotein-Cholesterol and Its Use as a Risk-Enhancer Test

Abstract

Background: Increased small dense low-density lipoprotein-cholesterol (sdLDL-C) is a risk factor for atherosclerotic cardiovascular disease (ASCVD) but typically requires advanced lipid testing. We describe two new equations, first one for calculating large buoyant LDL-C (lbLDL-C), based only upon results from the standard lipid panel, and the second one for sdLDL-C.

Methods: Equations for sdLDL-C and lbLDL-C were generated with least-squares regression analysis using the direct Denka sdLDL-C assay as reference (n = 20 171). sdLDL-C was assessed as a risk-enhancer test in the National Heart and Nutrition Examination Survey (NHANES), and for its association with ASCVD in the Multi-Ethnic Study of Atherosclerosis (MESA).

Results: The newly derived equations depend on two terms, namely LDL-C as determined by the Sampson equation, and an interaction term between LDL-C and the natural log of triglycerides (TG). The lbLDL-C equation (lbLDLC=1.43 × LDLC-0.14 ×(ln⁡(TG)× LDLC)- 8.99) was more accurate (R2 = 0.933, slope = 0.94) than the sdLDL-C equation (sdLDLC=LDLC- lbLDLC; R2 = 0.745, slope = 0.73). Using the 80th percentile (46 mg/dL) as a cut-point, sdLDL-C identified in NHANES additional high-risk patients not identified by other risk-enhancer tests based on TG, LDL-C, apolipoprotein B, and nonHDL-C. By univariate survival-curve analysis, estimated sdLDL-C was superior to other risk-enhancer tests in predicting ASCVD events in MESA. After multivariate adjustment for other known ASCVD risk factors, estimated sdLDL-C had the strongest association with ASCVD compared to other lipid parameters, including measured sdLDL-C.

Conclusions: Estimated sdLDL-C could potentially be calculated on all patients tested with a standard lipid panel to improve ASCVD risk stratification.

Fig. 1.

Distribution of cholesterol on different lipoprotein fractions. Cholesterol as measured by the direct Denka assay in lbLDL-C (A) or sdLDL-C (B) was plotted against LDL-C (n = 20 171). (C) Measured sdLDL-C (Msd-LDL-C) plotted against the natural log of TG (as a point of reference ln(150) = 5). Points on graphs are color-coded according to their TG or LDL-C percentile level as indicated in legend. (D) Cholesterol on Mlb-LDL-C (blue), Msd-LDL-C (red) and TRL-C (green) at various TG intervals.

Fig. 2.

Linear regression equations for estimating lbLDL-C and sdLDL-C. Linear regression equation for estimating lbLDL-C (Elb-LDL-C) (A) and sdLDL-C (Esd-LDL-C) (B). Dotted line represents the line of identity. Solid line is the linear fit for the indicated regression equation. Listed coefficient of determination (R²) and mean absolute deviation (MAD) values are from the validation dataset (n = 10 086), and the numbers in parentheses are corresponding values for the training dataset (n = 10 085). Residual error plots are shown for lbLDL-C (C, E, G) and sdLDL-C (D, F, H) for TG (C, D), LDL-C (E, F), and HDL-C (G, H). Points are color-coded according to the nonHDL-C percentile level as indicated in legend in Panel A.

Fig. 3.

Evaluation of Esd-LDL-C as a risk-enhancer test. (A) Elb-LDL-C was plotted against Esd-LDL-C by their percentile in the NHANES population without exclusion criteria (n = 13 085). Points are color-coded by their risk-enhancer group (see legend) as determined by the following risk-enhancer rules (TG > 175 mg/dL (points below blue line), LDL-C > 160 mg/dL (points above green line), and sdLDL-C > 46 mg/dL (points to the right of red line). (B) Venn diagram showing overlap for identified high-risk individuals based on the above risk-enhancer rules and risk-enhancer rules for apoB (>130 mg/dL) and nonHDL-C > 190 mg/dL). Numbers in parenthesis next to name of risk-enhancer rule show the percentage of total NHANES population that exceed the particular risk-enhancer rule. Numbers within Venn diagram show the percentage of NHANES population that fall within regions that overlap between the various risk-enhancer rules. (C) Mean value for each indicated lipid test is shown for groups color-coded as defined in legends. (D) Mean apoB/LDL-C (in mg/dL) ratio for indicated groups. (E) Mean 10-year ASCVD risk score is shown for each group defined in panel A, using the 4 pool-cohort risk equations for nonHispanic White females (NHWF), African American females (AAF), for nonHispanic White males (NHWM), and African American males (AAM) for age 65, with a systolic blood pressure of 120 mm Hg and no other risk factors. Lower case letters in Panels C–E above each bar indicates statistical difference between each group as determined by ANOVA. Means are ranked alphabetically with “a” being the highest. Shared letters between groups indicate no statistical difference (P > 0.05).

Fig. 4.

ASCVD survival curve analysis of risk-enhancer rules. Individuals in MESA, who had an sdLDL-C test performed by the direct Denka assay (n = 4610), were divided into two groups by whether they had results above (blue) or below (red) the 80th percentile for the indicated risk-enhancer rule cut-points. # failed indicates number of individuals who had an ASCVD event. Scores with an asterisk are statistically significant.