Advertisement

Atherosclerotic cardiovascular disease risk and small dense low-density lipoprotein cholesterol in men, women, African Americans and non-African Americans: The pooling project

  • Author Footnotes
    1 Joint first authorship since both authors contributed equally to this investigation.
    Ernst J. Schaefer
    Correspondence
    Corresponding author. Boston Heart Diagnostics, 200 Crossing Boulevard, Framingham, MA, 01702, USA.
    Footnotes
    1 Joint first authorship since both authors contributed equally to this investigation.
    Affiliations
    Cardiovascular Nutrition Laboratory, Human Nutrition Research Center on Aging at Tufts University, Department of Medicine, Tufts University School of Medicine, Friedman School of Nutrition Science and Policy at Tufts University, Boston, MA, USA

    Boston Heart Diagnostics, Framingham, MA, USA
    Search for articles by this author
  • Author Footnotes
    1 Joint first authorship since both authors contributed equally to this investigation.
    ,
    Author Footnotes
    2 Department of General Internal Medicine, Kyushu University Hospital, Fukuoka, Japan.
    Hiroaki Ikezaki
    Footnotes
    1 Joint first authorship since both authors contributed equally to this investigation.
    2 Department of General Internal Medicine, Kyushu University Hospital, Fukuoka, Japan.
    Affiliations
    Cardiovascular Nutrition Laboratory, Human Nutrition Research Center on Aging at Tufts University, Department of Medicine, Tufts University School of Medicine, Friedman School of Nutrition Science and Policy at Tufts University, Boston, MA, USA
    Search for articles by this author
  • Margaret R. Diffenderfer
    Affiliations
    Cardiovascular Nutrition Laboratory, Human Nutrition Research Center on Aging at Tufts University, Department of Medicine, Tufts University School of Medicine, Friedman School of Nutrition Science and Policy at Tufts University, Boston, MA, USA

    Boston Heart Diagnostics, Framingham, MA, USA
    Search for articles by this author
  • Elise Lim
    Affiliations
    Department of Biostatistics, Boston University School of Public Health, Boston, MA, USA

    The Framingham Heart Study, National Heart, Lung, Blood Institute, Framingham, MA, USA
    Search for articles by this author
  • Ching-Ti Liu
    Affiliations
    Department of Biostatistics, Boston University School of Public Health, Boston, MA, USA

    The Framingham Heart Study, National Heart, Lung, Blood Institute, Framingham, MA, USA
    Search for articles by this author
  • Ron C. Hoogeveen
    Affiliations
    Cardiovascular Research Section, Baylor College of Medicine, Houston, TX, USA
    Search for articles by this author
  • Weihua Guan
    Affiliations
    Department of Laboratory Medicine and Pathology, University of Minnesota School of Public Health, Minneapolis, MN, USA
    Search for articles by this author
  • Michael Y. Tsai
    Affiliations
    Department of Laboratory Medicine and Pathology, University of Minnesota School of Public Health, Minneapolis, MN, USA
    Search for articles by this author
  • Christie M. Ballantyne
    Affiliations
    Cardiovascular Research Section, Baylor College of Medicine, Houston, TX, USA

    Cardiology Division, Department of Medicine, Baylor College of Medicine, Houston, TX, USA
    Search for articles by this author
  • Author Footnotes
    1 Joint first authorship since both authors contributed equally to this investigation.
    2 Department of General Internal Medicine, Kyushu University Hospital, Fukuoka, Japan.
Open AccessPublished:January 18, 2023DOI:https://doi.org/10.1016/j.atherosclerosis.2023.01.015

      HIGHLIGHTS

      • An elevated direct sdLDL-C value > 50 mg/dL is an independent atherosclerotic cardiovascular disease (ASCVD) risk factor.
      • An elevated calculated sdLDL-C is not an independent ASCVD risk factor.
      • An elevated direct sdLDL-C value > 50 mg/dL is an ASCVD risk-enhancer.

      Abstract

      Background and aims

      Elevated small dense low-density lipoprotein-cholesterol (sdLDL-C) has been reported to be associated with increased atherosclerotic cardiovascular disease (ASCVD) risk. Our aims were to determine whether direct and calculated sdLDL-C were significant independent ASCVD risk factors in sex and race subgroups.

      Methods

      In a total of 15,933 participants free of ASCVD at baseline (median age 62 years, 56.7% female, 19.7% African American) fasting plasma lipids and sdLDL-C were measured by standardized automated methods. All subjects were followed for 10 years for incident ASCVD, which developed in 9.7% of subjects. SdLDL-C values were also calculated. Univariate and multivariate analyses were carried out to assess for independent associations with incident ASCVD after adjustment for all standard risk factors.

      Results

      All standard risk factors were significantly associated with incident ASCVD on univariate analysis, as were direct and calculated sdLDL-C. These latter parameters were also significant when added to the pooled cohort risk equation. However, associations were significantly stronger for direct sdLDL-C and were not significant for calculated values once direct values were in the model. In contrast to calculated values, top quartile direct sdLDL-C was significantly independently associated with incident ASCVD versus bottom quartile values in all subjects and subgroups, except African Americans (hazard ratios ≥1.50, p < 0.01). Subjects with direct values ≥ 50 mg/dL versus <25 mg/dL had significantly higher independent ASCVD risk in all groups (hazard ratios >1.49, all p < 0.01).

      Conclusions

      Having a direct small dense low-density lipoprotein cholesterol value ≥ 50 mg/dL is a significant independent ASCVD risk-enhancer.

      Graphical abstract

      Keywords

      1. Introduction

      Atherosclerotic cardiovascular disease (ASCVD) is a leading cause of death and disability. Established risk factors include sex, age, race, total cholesterol, high-density lipoprotein cholesterol (HDL-C), systolic blood pressure, blood pressure treatment, diabetes, and current smoking [
      • Goff Jr., D.C.
      • Lloyd-Jones D.M.
      • Bennett G.
      • Coady S.
      • D'Agostino R.B.
      • et al.
      American College of cardiology/American heart association task force on practice guidelines. 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of cardiology/American heart association task force on practice guidelines.
      ]. It has been recommended that patients be considered for statin therapy in addition to lifestyle modification if 1) their 10-year ASCVD risk is ≥ 7.5% based on the pooled cohort equation (PCE), 2) they have established ASCVD, 3) they have diabetes (between 40 and 75 years of age), or 4) they have a serum or plasma low-density lipoprotein cholesterol (LDL-C) level ≥190 mg/dL [
      • Goff Jr., D.C.
      • Lloyd-Jones D.M.
      • Bennett G.
      • Coady S.
      • D'Agostino R.B.
      • et al.
      American College of cardiology/American heart association task force on practice guidelines. 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of cardiology/American heart association task force on practice guidelines.
      ,
      • Grundy S.M.
      • Stone N.J.
      • Bailey A.L.
      • Beam C.
      • Birtcher K.K.
      • et al.
      AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/ADA/APHA/ASPC/NLA/PCNA guideline on the management of blood cholesterol: a report of the American College of cardiology/American heart association task force on clinical practice guidelines.
      ]. Additional biochemical risk enhancing factors listed by the recent guidelines panel included high-sensitivity C reactive protein ≥2.0 mg/L, lipoprotein(a) ≥50 mg/dL or ≥125 nmol/L, and/or triglycerides (TG) ≥175 mg/dL [
      • Grundy S.M.
      • Stone N.J.
      • Bailey A.L.
      • Beam C.
      • Birtcher K.K.
      • et al.
      AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/ADA/APHA/ASPC/NLA/PCNA guideline on the management of blood cholesterol: a report of the American College of cardiology/American heart association task force on clinical practice guidelines.
      ]. It has been documented that LDL-C-lowering with lifestyle modification and medications results in significant reductions in ASCVD morbidity and mortality [
      • Grundy S.M.
      • Stone N.J.
      • Bailey A.L.
      • Beam C.
      • Birtcher K.K.
      • et al.
      AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/ADA/APHA/ASPC/NLA/PCNA guideline on the management of blood cholesterol: a report of the American College of cardiology/American heart association task force on clinical practice guidelines.
      ,
      • Silverman M.G.
      • Ference B.A.
      • Im K.
      • Wiviott S.D.
      • Giugliano R.P.
      • et al.
      Association between lowering LDL-C and cardiovascular risk reduction among different therapeutic interventions: a systematic review and meta-analysis.
      ].
      Prior studies have shown that patients with ASCVD are significantly more likely to have higher small dense LDL and LDL particle numbers as compared to control subjects [
      • Schaefer E.J.
      • Tsunoda F.
      • Diffenderfer M.R.
      • Polisecki E.A.
      • Thai N.
      • et al.
      The measurement of lipids, lipoproteins, apolipoproteins, fatty acids, and sterols, and next generation sequencing for the diagnosis and treatment of lipid disorders.
      ,
      • Austin M.A.
      • King M.C.
      • Vranizan K.M.
      • Krauss R.M.
      Atherogenic lipoprotein phenotype. A proposed genetic marker for coronary heart disease risk.
      ,
      • Campos H.
      • Genest J.J.
      • Blijlevens E.
      • McNamara J.R.
      • Jenner J.
      • et al.
      Low-density lipoprotein particle size and coronary artery disease.
      ,
      • Mora S.
      • Szklo M.
      • Otvos J.D.
      • Greenland P.
      • Psaty B.M.
      • et al.
      LDL particle subclasses, LDL particle size, and carotid atherosclerosis in the Multi-Ethnic Study of Atherosclerosis (MESA).
      ,
      • Mora S.
      • Otvos J.D.
      • Rifai N.
      • Rosenson R.S.
      • Buring J.E.
      • et al.
      Lipoprotein particle profiles by nuclear magnetic resonance compared with standard lipids and apolipoproteins in predicting incident cardiovascular disease in women.
      ,
      • Mora S.
      • Caulfield M.P.
      • Wohlgemuth J.
      • Chen Z.
      • Superko H.R.
      • et al.
      Atherogenic lipoprotein subfractions determined by ion mobility and first cardiovascular events after random allocation to high-intensity statin or placebo: the JUPITER Trial.
      ]. However, assessments of lipoproteins using ultracentrifugation, gel electrophoresis, or nuclear magnetic resonance require special instrumentation or are labor-intensive. As a result, most laboratories measure only total cholesterol, TG, and HDL-C [
      • Burstein M.
      • Scholnick H.R.
      • Morfin R.
      Rapid method for the isolation of lipoproteins from human serum by precipitation with polyanions.
      ,
      ,
      • McNamara J.R.
      • Schaefer E.J.
      Automated enzymatic standardized lipid analyses for plasma and lipoprotein fractions.
      ,
      • Sugiuchi H.
      • Uji Y.
      • Okabe H.
      • Irie T.
      • Uekama K.
      • et al.
      Direct measurement of high-density lipoprotein cholesterol in serum with polyethylene glycol-modified enzymes and sulfated α-cyclodextrin.
      ]. Such assays were used in the currently recommended PCE [
      • Goff Jr., D.C.
      • Lloyd-Jones D.M.
      • Bennett G.
      • Coady S.
      • D'Agostino R.B.
      • et al.
      American College of cardiology/American heart association task force on practice guidelines. 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of cardiology/American heart association task force on practice guidelines.
      ]. Many laboratories use the Friedewald formula to calculate LDL-C from total cholesterol, TG, and HDL-C measured after overnight fasting [
      • Friedewald W.T.
      • Levy R.I.
      • Fredrickson D.S.
      Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge.
      ]. Other formulas have also been developed to calculate LDL-C [
      • Martin S.S.
      • Blaha M.J.
      • Elshazly M.B.
      • Toth P.P.
      • Kwiterovich P.O.
      • et al.
      Comparison of a novel method vs the Friedewald equation for estimating low-density lipoprotein cholesterol levels from the standard lipid profile.
      ,
      • Meeusen J.W.
      • Lueke A.J.
      • Jaffe A.S.
      • Saenger A.K.
      Validation of a proposed novel equation for estimating LDL cholesterol.
      ,
      • Sampson M.
      • Ling C.
      • Sun Q.
      • Harb R.
      • Ashmaig M.
      • et al.
      A new equation for calculation of low-density lipoprotein cholesterol in patients with normolipemia and/or hypertriglyceridemia.
      ]. Methods for directly measuring LDL-C have also been developed, as has a method for directly measuring small dense LDL cholesterol (sdLDL-C) [
      • McNamara J.R.
      • Cole T.G.
      • Contois J.H.
      • Ferguson C.A.
      • Ordovas J.M.
      • et al.
      Immunoseparation method for measuring low-density lipoprotein cholesterol directly from serum evaluated.
      ,
      • Okada M.
      • Matsui H.
      • Ito Y.
      • Fujiwara A.
      • Inano K.
      Low density lipoprotein cholesterol can be chemically measured; a new superior method.
      ,
      • Sakaue T.
      • Hirano T.
      • Yoshino G.
      • Sakai K.
      • Takeuchi H.
      • et al.
      Reactions of direct LDL-cholesterol assays with pure LDL fraction and IDL: comparison of three homogeneous methods.
      ,
      • Otokozawa S.
      • Ai M.
      • Asztalos B.F.
      • White C.C.
      • Demissie-Banjaw S.
      • et al.
      Direct assessment of plasma low-density lipoprotein and high-density lipoprotein cholesterol and coronary heart disease: results from the Framingham Offspring Study.
      ,
      • Hirano T.
      • Ito Y.
      • Saegusa H.
      • Yoshino G.
      A novel and simple method for quantification of small dense LDL.
      ,
      • Ito Y.
      • Fujimura M.
      • Ohta M.
      • Hirano T.
      Development of a homogeneous assay for measurement of small dense LDL cholesterol.
      ]. Elevated direct sdLDL-C values have been associated with increased ASCVD risk in prospective population studies [
      • Hirano T.
      • Ito Y.
      • Koba S.
      • Toyoda Ikejiri A.
      • et al.
      Clinical significance of small dense low-density lipoprotein cholesterol levels determined by the simple precipitation method.
      ,
      • Koba S.
      • Yokota Y.
      • Hirano T.
      • Ito Y.
      • Ban Y.
      • et al.
      Small LDL-cholesterol is superior to LDL-cholesterol for determining severe coronary atherosclerosis.
      ,
      • Ai M.
      • Otokozawa S.
      • Asztalos B.F.
      • Ito Y.
      • Nakajima K.
      • et al.
      Small dense LDL cholesterol and coronary heart disease: results from the Framingham Offspring Study.
      ,
      • Nishikura T.
      • Koba S.
      • Yokota Y.
      • Hirano T.
      • Tsunoda F.
      • et al.
      Elevated small dense low-density lipoprotein cholesterol as a predictor for future cardiovascular events in patients with stable coronary artery disease.
      ,
      • Arai H.
      • Kokubo Y.
      • Watanabe M.
      • Sawamura T.
      • Ito Y.
      • et al.
      Small dense low-density lipoproteins cholesterol can predict incident cardiovascular disease in an urban Japanese cohort: the Suita study.
      ,
      • Tsai M.Y.
      • Steffen B.T.
      • Guan W.
      • McClelland R.L.
      • Warnick R.
      • et al.
      New automated assay of small dense low-density lipoprotein cholesterol identifies risk of coronary heart disease: the Multi-ethnic Study of Atherosclerosis.
      ,
      • Hoogeveen R.C.
      • Gaubatz J.W.
      • Sun W.
      • Dodge R.C.
      • Crosby J.R.
      • et al.
      Small dense low-density lipoprotein-cholesterol concentrations predict risk for coronary heart disease: the Atherosclerosis Risk in Communities (ARIC) study.
      ,
      • Balling M.
      • Nordestgaard B.G.
      • Langsted A.
      • Varbo A.
      • Kamstrup P.R.
      • et al.
      Small dense low-density lipoprotein cholesterol predicts atherosclerotic cardiovascular disease in the Copenhagen General Population Study.
      ,
      • Ikezaki H.
      • Lim E.
      • Cupples L.A.
      • Liu C.T.
      • Asztalos B.F.
      • et al.
      Small dense low‐density lipoprotein cholesterol is the most atherogenic lipoprotein parameter in the prospective Framingham Offspring Study.
      ]. We have reported in the Framingham Offspring Study (FOS) that elevated sdLDL-C was the most atherogenic lipoprotein parameter, compared to direct LDL-C, lipoprotein(a), and other lipoprotein parameters, and was an independent predictor of incident ASCVD [
      • Ikezaki H.
      • Lim E.
      • Cupples L.A.
      • Liu C.T.
      • Asztalos B.F.
      • et al.
      Small dense low‐density lipoprotein cholesterol is the most atherogenic lipoprotein parameter in the prospective Framingham Offspring Study.
      ]. Our goals in this study were to test the hypotheses that directly measured sdLDL-C was an independent risk factor for ASCVD, coronary heart disease (CHD), and ischemic stroke in addition to the PCE and all standard ASCVD risk factors in men, women, non-African Americans, and African Americans. We analyzed a pooled sample from the Atherosclerosis Risk in Communities Study (ARIC), FOS, and the Multi-Ethnic Study of Atherosclerosis (MESA). We also compared our results with a recently published method to calculate sdLDL [
      • Sampson M.
      • Ling C.
      • Sun Q.
      • Harb R.
      • Ashmaig M.
      • et al.
      A new equation for calculation of low-density lipoprotein cholesterol in patients with normolipemia and/or hypertriglyceridemia.
      ,
      • Sampson M.
      • Wolska A.
      • Warnick R.
      • Lucero D.
      • Remaley A.T.
      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.
      ].

      2. Patients and methods

      2.1 Study design and study population

      We followed the “Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines in this study. The study meets the criteria of a “Cohort Study.” Study samples were collected from ARIC, FOS, and MESA participants, as previously described [
      • Tsai M.Y.
      • Steffen B.T.
      • Guan W.
      • McClelland R.L.
      • Warnick R.
      • et al.
      New automated assay of small dense low-density lipoprotein cholesterol identifies risk of coronary heart disease: the Multi-ethnic Study of Atherosclerosis.
      ,
      • Hoogeveen R.C.
      • Gaubatz J.W.
      • Sun W.
      • Dodge R.C.
      • Crosby J.R.
      • et al.
      Small dense low-density lipoprotein-cholesterol concentrations predict risk for coronary heart disease: the Atherosclerosis Risk in Communities (ARIC) study.
      ,
      • Ikezaki H.
      • Lim E.
      • Cupples L.A.
      • Liu C.T.
      • Asztalos B.F.
      • et al.
      Small dense low‐density lipoprotein cholesterol is the most atherogenic lipoprotein parameter in the prospective Framingham Offspring Study.
      ]. Information about all study participants is shown in Table 1. Most of the non-African American subjects were of European ancestry (Supplemental Table 1). A total of 15,933 participants (median age 62 years, 56.7% female, 19.7% African American, free of ASCVD at baseline) were followed for 10 years for incident ASCVD, which occurred in 9.7% of subjects as previously described (Supplemental Tables 2 and 3) [
      • Tsai M.Y.
      • Steffen B.T.
      • Guan W.
      • McClelland R.L.
      • Warnick R.
      • et al.
      New automated assay of small dense low-density lipoprotein cholesterol identifies risk of coronary heart disease: the Multi-ethnic Study of Atherosclerosis.
      ,
      • Hoogeveen R.C.
      • Gaubatz J.W.
      • Sun W.
      • Dodge R.C.
      • Crosby J.R.
      • et al.
      Small dense low-density lipoprotein-cholesterol concentrations predict risk for coronary heart disease: the Atherosclerosis Risk in Communities (ARIC) study.
      ,
      • Ikezaki H.
      • Lim E.
      • Cupples L.A.
      • Liu C.T.
      • Asztalos B.F.
      • et al.
      Small dense low‐density lipoprotein cholesterol is the most atherogenic lipoprotein parameter in the prospective Framingham Offspring Study.
      ]. All subjects were required to 1) be free of ASCVD at baseline, 2) have had plasma total cholesterol, TG, HDL-C, and direct sdLDL-C measured on frozen, previously unthawed, plasma from blood sampled after an overnight fast, 3) have had a baseline history and physical examination (including measurement of blood pressure, height, and weight) as part of their participation in the study, and 4) have follow-up clinical data available.
      Table 1Characteristics of all subjects at baseline by ASCVD event outcome.
      No Event(n = 14,385)ASCVD Event(n = 1548)HR (95% CI)p
      Demographics
       Women, n (%)8470 (58.9)568 (36.7)0.41 (0.37–0.45)2.44 × 10−65
       Age, year61 (56–67)65 (59–69)1.87 (1.73–2.01)5.31 × 10−60
       AA, n (%)2839 (19.7)302 (19.5)1.01 (0.89–1.15)0.869
       Non-AA, n (%)
      See Supplemental Table 1 for the ethnic demographics of non-African Americans.
      See Supplemental Table 1 for the ethnic demographics of non-African Americans.
      11,546 (80.3)1246 (80.5)0.99 (0.87–1.12)0.869
      Clinical/Treatment
       Systolic BP, mmHg123 (112–137)130 (118–145)1.52 (1.43–1.61)2.08 × 10−45
       BP Rx, n (%)4829 (33.6)750 (48.4)1.89 (1.71–2.09)7.12 × 10−36
       Diabetes, n (%)1516 (10.5)303 (19.6)2.17 (1.92–2.47)8.77 × 10−34
       Diabetes Rx, n (%)787 (5.5)177 (11.4)2.44 (2.09–2.86)8.44 × 10−29
       Cholesterol-lowering Rx, n (%)900 (6.3)169 (10.9)1.78 (1.52–2.09)1.26 × 10−12
       Smoking, n (%)2012 (14.0)287 (18.5)1.46 (1.29–1.66)5.98 × 10−9
      Lipids
       Total cholesterol, mg/dL199 (177–223)201 (178–225)1.06 (1.10–1.13)0.059
       Triglycerides, mg/dL115 (82–163)133 (94–185)1.12 (1.09–1.16)2.27 × 10−15
       LDL-C, (F), mg/dL
      Calculated using the Friedewald formula: (total cholesterol – HDL-C) – (triglycerides/5).
      121 (101–143)126 (105–148)1.19 (1.12–1.27)2.76 × 10−8
       LDL-C, (S), mg/dL17,
      Calculated using the Sampson formula [17].
      141 (121–163)146 (125–169)1.19 (1.12–1.26)1.54 × 10−8
       HDL-C, mg/dL49 (40–61)43 (36–53)0.59 (0.55–0.64)6.49 × 10−44
       Non-HDL-C, mg/dL
      Calculated using the following formula: total cholesterol – HDL-C.
      147 (124–172)156 (132–180)1.27 (1.20–1.35)2.15 × 10−15
       sdLDL-C, direct, mg/dL38.4 (27.7–53.1)44.7 (32.2–59.7)1.34 (1.27–1.42)9.87 × 10−24
       sdLDL-C, (S), mg/dL
      • Sampson M.
      • Wolska A.
      • Warnick R.
      • Lucero D.
      • Remaley A.T.
      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.
      ,
      Calculated using the Sampson formula [33].
      36.6 (29.0–46.2)40.3 (31.6–48.3)1.34 (1.27–1.43)1.27 × 10−23
       lbLDL-C (S), mg/dL
      • Sampson M.
      • Ling C.
      • Sun Q.
      • Harb R.
      • Ashmaig M.
      • et al.
      A new equation for calculation of low-density lipoprotein cholesterol in patients with normolipemia and/or hypertriglyceridemia.
      ,
      • Sampson M.
      • Wolska A.
      • Warnick R.
      • Lucero D.
      • Remaley A.T.
      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.
      ,
      Calculated using the Sampson formulas [17,33].
      99 (83–115)100 (84–116)1.06 (0.99–1.12)0.085
      Values are median and interquartile range (25th‒75th percentile) for continuous variables or number (%) for categorical variables. Hazard ratios for continuous variables represent comparison of risk across the interquartile range with the No Event as reference. Variables not normally distributed were log-transformed prior to regression analysis. To convert cholesterol from mg/dL to mmol/L, multiply by 0.02586; to convert triglyceride from mg/dL to mmol/L, multiple by 0.01129.
      AA, African American/black; ASCVD, atherosclerotic cardiovascular disease; BP, blood pressure; CI, confidence interval; HDL-C, high-density lipoprotein cholesterol; HR, hazard ratio; lbLDL-C, large buoyant low-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; no event, subjects who had no event or procedure prior to baseline and no event or procedure during 10-year follow-up; Rx, treatment; sdLDL-C, small dense low-density lipoprotein cholesterol.
      a See Supplemental Table 1 for the ethnic demographics of non-African Americans.
      b Calculated using the Friedewald formula: (total cholesterol – HDL-C) – (triglycerides/5).
      c Calculated using the Sampson formula [
      • Sampson M.
      • Ling C.
      • Sun Q.
      • Harb R.
      • Ashmaig M.
      • et al.
      A new equation for calculation of low-density lipoprotein cholesterol in patients with normolipemia and/or hypertriglyceridemia.
      ].
      d Calculated using the following formula: total cholesterol – HDL-C.
      e Calculated using the Sampson formula [
      • Sampson M.
      • Wolska A.
      • Warnick R.
      • Lucero D.
      • Remaley A.T.
      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.
      ].
      f Calculated using the Sampson formulas [
      • Sampson M.
      • Ling C.
      • Sun Q.
      • Harb R.
      • Ashmaig M.
      • et al.
      A new equation for calculation of low-density lipoprotein cholesterol in patients with normolipemia and/or hypertriglyceridemia.
      ,
      • Sampson M.
      • Wolska A.
      • Warnick R.
      • Lucero D.
      • Remaley A.T.
      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.
      ].
      All subjects provided information about their medical history and use of medications and supplements. Hypertension was defined as a blood pressure of >140/90 mmHg or being on medications for hypertension. Diabetes was defined as a fasting glucose ≥125 mg/dL or being on medications for diabetes. Smoking was defined as cigarette smoking within the past year. At baseline, 6.7% of the subjects were taking lipid-lowering medications and were included in the study sample (Table 1).
      All studies were carried out with the approval of the human institutional review boards at the respective institutions and conform to the ethical guidelines of the 1975 Declaration of Helsinki, as previously described [
      • Tsai M.Y.
      • Steffen B.T.
      • Guan W.
      • McClelland R.L.
      • Warnick R.
      • et al.
      New automated assay of small dense low-density lipoprotein cholesterol identifies risk of coronary heart disease: the Multi-ethnic Study of Atherosclerosis.
      ,
      • Hoogeveen R.C.
      • Gaubatz J.W.
      • Sun W.
      • Dodge R.C.
      • Crosby J.R.
      • et al.
      Small dense low-density lipoprotein-cholesterol concentrations predict risk for coronary heart disease: the Atherosclerosis Risk in Communities (ARIC) study.
      ,
      • Ikezaki H.
      • Lim E.
      • Cupples L.A.
      • Liu C.T.
      • Asztalos B.F.
      • et al.
      Small dense low‐density lipoprotein cholesterol is the most atherogenic lipoprotein parameter in the prospective Framingham Offspring Study.
      ]. Written informed consent was obtained from all study participants. The research committees of all studies provided approval for these pooled analyses to be carried out.

      2.2 Criteria for ASCVD, CHD, and stroke events

      The following criteria for ASCVD endpoints were used: the development of myocardial infarction (recognized with or without diagnostic electrocardiogram, but including enzymes and history or recognized at the time of autopsy), coronary revascularization (angioplasty or coronary artery bypass grafting) during the follow-up period, stroke (atherothrombotic brain infarction, cerebral embolism, definite or other cardiovascular accident, intracerebral hemorrhage or subarachnoid hemorrhage), and death from either myocardial infarction or stroke (sudden death from CHD; death from a cerebrovascular accident; death from other cardiovascular diseases). For CHD criteria, we used ASCVD criteria, except that all subjects that had a stroke were excluded. For ischemic stroke criteria, we excluded all subjects who developed CHD and included only those who had an ischemic stroke, excluding subjects with an intracerebral or a subarachnoid hemorrhage. Only the first event over a follow-up time of 10 years was used in the analysis.

      2.3 Laboratory measurements

      Fasting plasma samples stored at −80 °C and never thawed were used for the analysis. Total cholesterol, TG, and HDL-C were determined by standard enzymatic methods as previously described [
      • Otokozawa S.
      • Ai M.
      • Asztalos B.F.
      • White C.C.
      • Demissie-Banjaw S.
      • et al.
      Direct assessment of plasma low-density lipoprotein and high-density lipoprotein cholesterol and coronary heart disease: results from the Framingham Offspring Study.
      ,
      • Tsai M.Y.
      • Steffen B.T.
      • Guan W.
      • McClelland R.L.
      • Warnick R.
      • et al.
      New automated assay of small dense low-density lipoprotein cholesterol identifies risk of coronary heart disease: the Multi-ethnic Study of Atherosclerosis.
      ,
      • Hoogeveen R.C.
      • Gaubatz J.W.
      • Sun W.
      • Dodge R.C.
      • Crosby J.R.
      • et al.
      Small dense low-density lipoprotein-cholesterol concentrations predict risk for coronary heart disease: the Atherosclerosis Risk in Communities (ARIC) study.
      ,
      • Ikezaki H.
      • Lim E.
      • Cupples L.A.
      • Liu C.T.
      • Asztalos B.F.
      • et al.
      Small dense low‐density lipoprotein cholesterol is the most atherogenic lipoprotein parameter in the prospective Framingham Offspring Study.
      ]. Direct sdLDL-C was measured at the respective laboratories using assay kits from the Denka Corporation (Tokyo, Japan) as previously described [
      • Ito Y.
      • Fujimura M.
      • Ohta M.
      • Hirano T.
      Development of a homogeneous assay for measurement of small dense LDL cholesterol.
      ,
      • Tsai M.Y.
      • Steffen B.T.
      • Guan W.
      • McClelland R.L.
      • Warnick R.
      • et al.
      New automated assay of small dense low-density lipoprotein cholesterol identifies risk of coronary heart disease: the Multi-ethnic Study of Atherosclerosis.
      ,
      • Hoogeveen R.C.
      • Gaubatz J.W.
      • Sun W.
      • Dodge R.C.
      • Crosby J.R.
      • et al.
      Small dense low-density lipoprotein-cholesterol concentrations predict risk for coronary heart disease: the Atherosclerosis Risk in Communities (ARIC) study.
      ,
      • Ikezaki H.
      • Lim E.
      • Cupples L.A.
      • Liu C.T.
      • Asztalos B.F.
      • et al.
      Small dense low‐density lipoprotein cholesterol is the most atherogenic lipoprotein parameter in the prospective Framingham Offspring Study.
      ]. The direct sdLDL-c method uses a polyoxyethylene benzylphenyl ether derivative to dissociate triglyceride-rich lipoproteins and HDL. Sphingomyelinase is used for dissociation of large buoyant LDL (lbLDL) from LDL owing to the higher sphingomyelin content in the large LDL subfractions. A polyoxyethylene styrenephenyl ether derivative is used to protect sdLDL from the dissociative actions of sphingomyelinase. In addition, cholesterol oxidase/esterase is used during an initial incubation step. Thereafter, polyoxyethylene alkyl ether is used to dissociate the sdLDL; and the cholesterol released is measured using cholesterol oxidase/esterase reagent on high through-put chemistry analyzers. The results of this homogeneous method correlated very well with results obtained with ultracentrifugation for sdLDL (density 1.044–1.063 g/mL, r = 0.99), with within-run precision coefficients of <1.1% as described by Ito and colleagues [
      • Ito Y.
      • Fujimura M.
      • Ohta M.
      • Hirano T.
      Development of a homogeneous assay for measurement of small dense LDL cholesterol.
      ]. All analyses were run on automated, high-throughput analyzers with within- and between-run coefficients of variation of <5.0%. Stability of sdLDL-C for >5 years has been documented, provided the plasma or serum has been kept continuously at −80 °C and never thawed until use. The direct sdLDL-C assay used in these studies has been approved for ASCVD risk assessment by the United States Food and Drug Administration.
      To convert all cholesterol values from mg/dL to mmol/L, multiply by 0.02586. To convert TG values from mg/dL to mmol/L, multiple by 0.01129.

      2.4 Calculated biochemical variables

      Non-HDL-C, LDL-C, and sdLDL-C were calculated as follows:
      • 1)
        Non-HDL-C was calculated as total cholesterol (TC) – HDL-C.
      • 2)
        LDL-CF (Friedewald formula) was calculated as TC – HDL-C - TG/5 [
        • Friedewald W.T.
        • Levy R.I.
        • Fredrickson D.S.
        Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge.
        ].
      • 3)
        LDL-CS (Sampson formula) was calculated as TC/0.948 - HDL-C/0.971- (TG/8.56 + TG x Non-HDL-C/2140 – TG2/16,100) – 9.44 [
        • Sampson M.
        • Ling C.
        • Sun Q.
        • Harb R.
        • Ashmaig M.
        • et al.
        A new equation for calculation of low-density lipoprotein cholesterol in patients with normolipemia and/or hypertriglyceridemia.
        ].
      • 4)
        lbLDL-CS (Sampson formula) was calculated as 1.43 x calculated LDL-C – [0.14 x (log TG x LDL-C)] – 8.99 [
        • Sampson M.
        • Wolska A.
        • Warnick R.
        • Lucero D.
        • Remaley A.T.
        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.
        ].
      • 5)
        sdLDL-CS (Sampson formula) was calculated as calculated LDL-CS – lbLDL-C [
        • Sampson M.
        • Wolska A.
        • Warnick R.
        • Lucero D.
        • Remaley A.T.
        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.
        ].
      It is important to emphasize that in the ultimate calculation of sdLDL-C using the Sampson formula as above, TC, HDL-C, and non-HDL-C are each used once, while TG is used four times.

      2.5 Statistical analysis

      Information on all statistical analyses, using methods previously described [
      • Ikezaki H.
      • Lim E.
      • Cupples L.A.
      • Liu C.T.
      • Asztalos B.F.
      • et al.
      Small dense low‐density lipoprotein cholesterol is the most atherogenic lipoprotein parameter in the prospective Framingham Offspring Study.
      ,
      • Pencina M.J.
      • D'Agostino R.B.
      Evaluating discrimination of risk prediction models: the C Statistic.
      ,
      • Ikezaki H.
      • Fisher V.A.
      • Lim E.
      • Ai M.
      • Liu C.T.
      • et al.
      Direct versus calculated low-density lipoprotein cholesterol and C reactive protein in cardiovascular disease risk assessment in the Framingham Offspring Study.
      ], is found in the Supplemental Material. Data from all three studies were pooled and analyzed in a blinded fashion. Assessments of ASCVD, CHD, and ischemic stroke risk were carried out in an unadjusted fashion comparing median and interquartile range IQR) values between groups, and also by quartile values comparing the hazard ratio (HR) for each of the top three quartiles with the bottom quartile. Correlation coefficient analysis and Cox proportional hazard modeling for multivariate analysis were performed. We also used multivariate cutpoint analysis for direct sdLDL-C, after adjustment for all standard ASCVD risk factors, to determine whether direct sdLDL-C could serve as a significant risk enhancer.

      3. Results

      3.1 Univariate unadjusted analyses

      As shown, of 15,933 total subjects studied, 9.72% had an ASCVD event, 7.33% had a CHD event, 2.56% had a stroke, and 2.10% had an ischemic stroke over the 10-year follow-up period (Table 1; Supplemental Table 2). ASCVD, CHD, and ischemic stroke risks in women were 59%, 66%, and 43%, respectively, lower than those observed for men (all p < 0.0001) based on a comparison of hazard ratios (Table 1; Supplemental Tables 2–8). These hazard ratios were based on a direct comparison of incident rates for ASCVD observed over the follow-up period in men and women, respectively (Table 1). The risk for these same events was 87%, 77%, and 244%, respectively, higher in older subjects (>67.0 years) as compared to younger subjects (<56.0 years, all p < 0.0001) (Table 1; Supplemental Tables 2–8). African Americans in our study had a similar risk for ASCVD, a 17% lower risk for CHD (p < 0.05), and a 76% higher risk for ischemic stroke as compared to non-African Americans (Table 1; Supplemental Tables 2–8). Women had significantly lower sdLDL-C and higher HDL-C levels than men, and the same was true for African Americans versus non-African Americans (Supplemental Tables 2 and 3). African Americans had significantly higher systolic blood pressure, blood pressure treatment, diabetes, diabetes treatment, and smoking, and significantly less cholesterol treatment than non-African Americans (Supplemental Table 3). All standard risk factors and other parameters, including direct and calculated sdLDL-C, were significantly (p < 0.01) different at baseline between subjects who developed ASCVD, CHD, and stroke, except for total cholesterol and calculated lbLDL-C in some instances. For ischemic stroke, only associations with TG, HDL-C, direct sdLDL-C, and calculated sdLDL-C were significant (Table 1; Supplemental Tables 2–8).
      Unadjusted Kaplan Meier plots by quartiles of direct sdLDL-C indicated a higher risk for higher quartile values compared to bottom quartile values for incident ASCVD for all subjects and all subgroup analysis (Fig. 1). HR for all groups for top quartile direct sdLDL-C versus bottom quartile values ranged from 1.57 to 1.93 (all p < 0.0001). We assessed the intercorrelations of the measured biochemical factors (total cholesterol, TG, HDL-C, direct sdLDL-C) with the calculated values (calculated LDL-C, sdLDL-C, and non-HDL-C) (Supplemental Table 9). Not surprisingly, the correlation between calculated LDL-C using the Friedewald formula and the value calculated with the Sampson formula was very high at 0.996 (r2 = 0.992). Moreover, correlations between non-HDL-C and calculated LDL-C (Sampson) and between non-HDL-C and calculated sdLDL-C (Sampson) were also quite high (0.947 and 0.890, respectively). However, the correlation between direct sdLDL-C and calculated sdLDL-C (Sampson) was lower at 0.821 (r2 = 0.674). These data indicated to us that calculated sdLDL-C only accounted for about 67% of the variability in direct sdLDL-C and might not perform as well as direct sdLDL-C on multivariate analysis after adjustment for standard risk factors. The correlations between direct sdLDL-C and TG were 0.637 (Pearson) and 0.670 (Spearman), indicating that about 45% of sdLDL-C variability could be accounted for by variability in TG. The strongest parameter correlated with sdLDL-C was non-HDL-C at about 0.760 (both Pearson and Spearman correlation coefficients).
      Fig. 1
      Fig. 1Unadjusted Kaplan-Meier 10-year ASCVD event risk analysis by quartiles of direct sdLDL-C.
      (A) All subjects; (B) men; (C) women; (D) non-African Americans; (E) African Americans. Green line depicts 1st quartile; black line: 2nd quartile; red line: 3rd quartile; blue line: 4th quartile. Unadjusted HR compares all quartiles with the 1st quartile. Quartile concentration range and unadjusted HR (95% CI) are shown in and . Adjusted HR (95% CI) for each quartile for ASCVD risk is shown in . reports the ethnic demographics of the non-African American cohort.ASCVD, atherosclerotic cardiovascular disease; CI, confidence interval; HR, hazard ratio; sdLDL-C, small dense low-density lipoprotein cholesterol.

      3.2 Multivariate analysis

      3.2.1 Pooled cohort equation (PCE)

      We assessed the extent to which biochemical values provided incremental information about ASCVD risk for all subjects above and beyond that provided using the PCE model (Fig. 2) [
      • Goff Jr., D.C.
      • Lloyd-Jones D.M.
      • Bennett G.
      • Coady S.
      • D'Agostino R.B.
      • et al.
      American College of cardiology/American heart association task force on practice guidelines. 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of cardiology/American heart association task force on practice guidelines.
      ]. Statistically significant information about ASCVD risk on top of the PCE model was added by log direct sdLDL-C, log TG, log calculated sdLDL-C, calculated non-HDL-C, calculated LDL-C (Sampson), and calculated LDL-C (Friedewald). The strongest association was observed for log direct sdLDL-C (HR 1.27, 95% CI 1.18–1.36, p < 0.00001). Moreover, when log direct sdLDL-C was added to the PCE model, no other measured or calculated parameter added significant information about risk for ASCVD regardless of the order of parameter entry. When log TG was added to the PCE model first and then log sdLDL-C was added, only the latter parameter remained significant. The same was the case for all other multivariate models. The change noted for the model C statistic values when log direct sdLDL-C was added to the PCE model was statistically significant (p < 0.05); this was not the case for any other parameter. However, this value was 0.638 for ASCVD, indicating that the PCE model could be substantially improved, possibly by adjustment of the formula.
      Fig. 2
      Fig. 2Association with 10-year ASCVD event risk for all subjects when atherogenic biomarkers are added individually to the pooled cohort equation (PCE) model.
      The PCE model includes age, sex, race (African American or non-African American), total cholesterol, HDL-C, systolic blood pressure, antihypertensive medication usage, diabetic status, and smoking. Log TG, non-HDL-C, calculated LDL-C, and log calculated sdLDL-C added no significant information about ASCVD risk after log sdLDL-C was entered into the model, regardless of the order of parameters entered. The C statistic for ASCVD risk was significantly (p < 0.01) greater with sdLDL-C than without sdLDL-C (C statistic increased from 0.631 to 0.638). This was not the case for calculated sdlDL-C. See for the ethnic demographics of the non-African American cohort. aHRadj (95% CI) is expressed as the 10-year event risk for cases compared to non-cases when atherogenic biomarker was added to the PCE and is based on the interquartile range for each measured or calculated biochemical variable. bsdLDL-C, calculated sdLDL-C, and TG were not normally distributed and were log-transformed prior to statistical analysis. cCalculated using the Sampson formula [
      • Sampson M.
      • Wolska A.
      • Warnick R.
      • Lucero D.
      • Remaley A.T.
      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.
      ]. dCalculated using the Sampson formula [
      • Sampson M.
      • Ling C.
      • Sun Q.
      • Harb R.
      • Ashmaig M.
      • et al.
      A new equation for calculation of low-density lipoprotein cholesterol in patients with normolipemia and/or hypertriglyceridemia.
      ]. eCalculated using the Friedewald formula: (total cholesterol – HDL-C) – (TG/5). fCalculated using the Sampson formulas [
      • Sampson M.
      • Ling C.
      • Sun Q.
      • Harb R.
      • Ashmaig M.
      • et al.
      A new equation for calculation of low-density lipoprotein cholesterol in patients with normolipemia and/or hypertriglyceridemia.
      ,
      • Sampson M.
      • Wolska A.
      • Warnick R.
      • Lucero D.
      • Remaley A.T.
      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.
      ].ASCVD, atherosclerotic cardiovascular disease; calc, calculated; CI, confidence interval; HRadj, adjusted hazard ratio; lbLDL-C, large buoyant low-density lipoprotein cholesterol; LDL-C (F), low-density lipoprotein cholesterol calculated using Friedewald formula; LDL-C (S), low-density lipoprotein cholesterol calculated using Sampson formula; HDL-C, high-density lipoprotein cholesterol; PCE, pooled cohort equation; sdLDL-C, small dense low-density lipoprotein cholesterol; TG, triglycerides.

      3.2.2 Quartile analysis

      We assessed the extent to which quartiles of direct and calculated sdLDL-C added information about incident ASCVD risk after adjusting for all standard risk factors as well as for the use of cholesterol-lowering medication (Fig. 3). Men, women, and non-African American subjects with direct sdLDL-C values in quartiles 2–4 had a significantly higher ASCVD risk than those with values in the bottom quartile. The African American subjects did not; however, the trends for African American subjects were the same as for other groups. C statistic values ranged from 0.680 to 0.721 in the groups, substantially higher than what we observed with the PCE even after the addition of direct sdLDL-C. This finding was not the case for top quartile calculated sdLDL-C in any group, indicating that in contrast to direct sdLDL-C, calculated sdLDL-C values had no significant effect on ASCVD risk after adjustment for standard ASCVD risk factors.
      Fig. 3
      Fig. 3Association with 10-year ASCVD event risk across quartiles of direct and calculated sdLDL-C, adjusted for all risk factors.Graph, 10-year ASCVD event risk and direct sdLDL-C.
      Table, 10-year ASCVD event risk and direct sdLDL-C compared with calculated sdLDL-C (Sampson). HR (95% CI) is shown in comparison to the lowest quartile (Q1), after adjustment for age, sex, race, diabetes, hypertension, hypertension treatment, smoking, HDL-C, total cholesterol, and cholesterol-lowering medication. p value and C statistic in the graph represent trend across the quartiles of direct sdLDL-C. Calculated sdLDL-C (data shown), LDL-C calculated by either Friedewald or Sampson formula, and calculated lbLDL-C (data not shown) were not significant after adjustment. C statistic without sdLDL-C in all subjects, males, females, non-AA, and AA was 0.718, 0.676, 0.695, 0.719, and 0.715, respectively. Black squares represent HRadj for lowest quartile (Q1); light blue squares: HRadj for Q2; dark blue squares: HRadj for Q3; red squares: HRadj for Q4. Error bars show 95% CI. See for the ethnic demographics of the non-AA cohort. To convert sdLDL-C values from mg/dL to mmol/L, multiply by 0.02586.AA, African-American/black; C, C statistic; CI, confidence interval; HRadj, adjusted hazard ratio; Q, quartile; ref, reference; sdLDL-C, small dense low-density lipoprotein cholesterol.

      3.2.3 Cutpoint analysis

      In order to provide clinically useful information for clinicians, we also carried out multivariate analysis comparing direct sdLDL-C cutpoint values of ≥50 mg/dL (approximate top quartile value) with optimal values of <25 mg/dL (approximate bottom quartile value) after adjustment for all other risk factors including use of cholesterol-lowering medication. A direct sdLDL-C value ≥ 50 mg/dL was associated in all groups with statistically significant increased ASCVD risk (HR range 1.49–1.88; C statistic range 0.679–0.721 [Fig. 4]) and CHD risk (HR range 1.63–1.93; C statistic range 0.680–0.743 [Supplemental Fig. 1]), but not for ischemic stroke risk (HR 1.35–2.06, not significant but positive trends noted [Supplemental Fig. 2]). When TG was added to any multivariate model either before or after sdLDL-C, it was not significant when all other standard parameters were in the model.
      Fig. 4
      Fig. 4Association with 10-year ASCVD event risk across cutpoints of direct sdLDL-C, adjusted for all risk factors.
      Direct sdLDL-C values were classified according to prespecified clinical cutpoints. HRadj (95% CI), shown in comparison to the lowest cutpoint group, was adjusted for age, sex, race, diabetic status, hypertension, hypertension treatment, smoking, high-density lipoprotein cholesterol, total cholesterol, and cholesterol-lowering medication. p value and C statistic in the graph represent trend across the sdLDL-C cutpoints for each cohort. C statistic values without sdLDL-C for all subjects, men, women, non-AA, and AA were 0.718, 0.676, 0.695, 0.719, and 0.715, respectively (all p < 0.05 versus C statistic values with sdLDL-C). Black squares represent HRadj for sdLDL-C <25 mg/dL; dark blue squares: sdLDL-C 25.0-<50.0 mg/dL; red squares: sdLDL-C ≥50.0 mg/dL. Error bars show 95% CI. See for the ethnic demographics of the non-AA cohort. To convert sdLDL-C values from mg/dL to mmol/L, multiply by 0.02586.AA, African American; ASCVD, atherosclerotic cardiovascular disease; C, C statistic; CI, confidence interval; HRadj, adjusted hazard ratio; ref, reference; sdLDL-C, small dense low-density lipoprotein cholesterol.

      4. Discussion

      4.1 ASCVD risk, sex, age, and race

      ASCVD remains the leading cause of death in the United States. In this analysis for subjects with a median age of 62 years, about 10% had an ASCVD event over 10 years, about 7% had a CHD event, and about 2.5% had a stroke (about 2% had an ischemic stroke). As expected, the observed incidence for these events in women was about half of those observed for men in this age group. These differences were related to the lower smoking rates, higher HDL-C levels, and somewhat lower direct sdLDL-C levels observed in women compared to men. For older subjects, the observed incidence for these same endpoints was almost twice as high as in younger subjects and was substantially higher for strokes. These differences were related to higher prevalence rates of hypertension, diabetes, and hyperlipemia in the elderly compared to younger subjects. African Americans in our study had a similar observed incidence for ASCVD or CHD events, but a greatly increased observed incidence rate for ischemic stroke as compared to non-African Americans. These differences in stroke incidence rates were associated with higher systolic blood pressure, higher diabetes prevalence, and more smoking observed in African Americans than in non-African Americans.

      4.2 Small dense LDL-C and ASCVD risk

      Our study supports prior reports from Japan, the United States, and Denmark that elevated direct sdLDL-C is significantly related to prospective ASCVD and CHD risk [
      • Hirano T.
      • Ito Y.
      • Koba S.
      • Toyoda Ikejiri A.
      • et al.
      Clinical significance of small dense low-density lipoprotein cholesterol levels determined by the simple precipitation method.
      ,
      • Koba S.
      • Yokota Y.
      • Hirano T.
      • Ito Y.
      • Ban Y.
      • et al.
      Small LDL-cholesterol is superior to LDL-cholesterol for determining severe coronary atherosclerosis.
      ,
      • Ai M.
      • Otokozawa S.
      • Asztalos B.F.
      • Ito Y.
      • Nakajima K.
      • et al.
      Small dense LDL cholesterol and coronary heart disease: results from the Framingham Offspring Study.
      ,
      • Nishikura T.
      • Koba S.
      • Yokota Y.
      • Hirano T.
      • Tsunoda F.
      • et al.
      Elevated small dense low-density lipoprotein cholesterol as a predictor for future cardiovascular events in patients with stable coronary artery disease.
      ,
      • Arai H.
      • Kokubo Y.
      • Watanabe M.
      • Sawamura T.
      • Ito Y.
      • et al.
      Small dense low-density lipoproteins cholesterol can predict incident cardiovascular disease in an urban Japanese cohort: the Suita study.
      ,
      • Tsai M.Y.
      • Steffen B.T.
      • Guan W.
      • McClelland R.L.
      • Warnick R.
      • et al.
      New automated assay of small dense low-density lipoprotein cholesterol identifies risk of coronary heart disease: the Multi-ethnic Study of Atherosclerosis.
      ,
      • Hoogeveen R.C.
      • Gaubatz J.W.
      • Sun W.
      • Dodge R.C.
      • Crosby J.R.
      • et al.
      Small dense low-density lipoprotein-cholesterol concentrations predict risk for coronary heart disease: the Atherosclerosis Risk in Communities (ARIC) study.
      ,
      • Balling M.
      • Nordestgaard B.G.
      • Langsted A.
      • Varbo A.
      • Kamstrup P.R.
      • et al.
      Small dense low-density lipoprotein cholesterol predicts atherosclerotic cardiovascular disease in the Copenhagen General Population Study.
      ,
      • Ikezaki H.
      • Lim E.
      • Cupples L.A.
      • Liu C.T.
      • Asztalos B.F.
      • et al.
      Small dense low‐density lipoprotein cholesterol is the most atherogenic lipoprotein parameter in the prospective Framingham Offspring Study.
      ]. Moreover, our findings extend prior conclusions to indicate that sdLDL-C provides additional information about ASCVD risk even after controlling for all risk factors, including HDL-C, total cholesterol, and cholesterol-lowering medication [
      • Ikezaki H.
      • Lim E.
      • Cupples L.A.
      • Liu C.T.
      • Asztalos B.F.
      • et al.
      Small dense low‐density lipoprotein cholesterol is the most atherogenic lipoprotein parameter in the prospective Framingham Offspring Study.
      ]. We observed that the direct sdLDL-C relationship with ASCVD risk was incremental, increasing with each quartile of direct sdLDL-C concentration. We also documented the significance of direct sdLDL-C as an ASCVD risk factor using univariate analysis and using multivariate analysis after adjustment for the PCE or after adjustment for all other risk factors using either quartile or cutpoint analysis. These results indicate that direct sdLDL-C is a significant independent risk-enhancer for ASCVD. Importantly for the clinician, our data indicate that a direct sdLDL-C value ≥ 50 mg/dL is a significant risk-enhancer for ASCVD after controlling for all standard risk factors. It is well known that the five criteria used for the metabolic syndrome (diabetes, hypertension, hypertriglyceridemia, elevated body mass index, and low HDL-C) are all correlated with one another and with direct sdLDL-C. However, in our analyses, diabetes, hypertension, low HDL-C, and elevated sdLDL-C were all independent ASCVD risk factors; but this was not the case for TG.
      Because of recent publications citing the benefits of using calculated LDL-C and calculated sdLDL-C, we evaluated these parameters in our pooled cohort [
      • Sampson M.
      • Ling C.
      • Sun Q.
      • Harb R.
      • Ashmaig M.
      • et al.
      A new equation for calculation of low-density lipoprotein cholesterol in patients with normolipemia and/or hypertriglyceridemia.
      ,
      • Sampson M.
      • Wolska A.
      • Warnick R.
      • Lucero D.
      • Remaley A.T.
      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.
      ]. Because direct LDL-C was not measured in ARIC and MESA, we could not perform a direct comparison with the new method for calculated LDL-C for our entire cohort. In FOS where direct LDL-C was measured, we have documented previously that direct LDL-C added significant information about ASCVD risk to the PCE model and on multivariate analysis, while calculated LDL-C (Friedewald and Martin formulas) did not [
      • Ikezaki H.
      • Fisher V.A.
      • Lim E.
      • Ai M.
      • Liu C.T.
      • et al.
      Direct versus calculated low-density lipoprotein cholesterol and C reactive protein in cardiovascular disease risk assessment in the Framingham Offspring Study.
      ]. The new calculated LDL-C (Sampson) values performed well on univariate analysis but did not add significant information with regard to ASCVD risk on top of the PCE model or on multivariate analysis once direct sdLDL-C was in the model. In this study, we documented that calculated LDL-C (Sampson) values were highly correlated with LDL-C (Friedewald) (r > 0.99) (Supplemental Table 9). In addition, we documented that direct sdLDL-C added significant information about ASCVD risk on top of the PCE model and using quartile analysis after adjustment for all ASCVD risk factors. This was not the case for calculated sdLDL-C. The direct sdLDL-C assay can be run on various automated high-throughput analyzers, has been approved by the Food and Drug Administration, and is available from reference laboratories in the United States and in other countries.
      In the current cholesterol guidelines, it has been recommended that values of TG ≥ 175 mg/dL, high-sensitivity C-reactive protein ≥2.0 mg/L, and lipoprotein (a) ≥50 mg/dL or ≥125 nmol/L be considered as ASCVD risk-enhancing factors [
      • Grundy S.M.
      • Stone N.J.
      • Bailey A.L.
      • Beam C.
      • Birtcher K.K.
      • et al.
      AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/ADA/APHA/ASPC/NLA/PCNA guideline on the management of blood cholesterol: a report of the American College of cardiology/American heart association task force on clinical practice guidelines.
      ]. The presence of such risk-enhancing factors may favor statin therapy in patients with borderline 10-year ASCVD risk. In this study we assessed the clinical utility of sdLDL-C levels and documented that a sdLDL-C of ≥50 mg/dL should be considered as an ASCVD risk-enhancing factor. An elevated value increases ASCVD risk by about 50% on top of established risk factors in all subject groups as compared to those with values < 25 mg/dL. Moreover, for both ASCVD and CHD risk, elevated direct sdLDL-C significantly improved risk assessment both on top of the PCE model, by quartile analysis and by cut-point analysis. In our view, sdLDL is the most atherogenic lipoprotein because its prolonged residence time allows for a greater probability of modification and oxidation, and its smaller size allows for greater penetration into the arterial wall [
      • Thongtang N.
      • Diffenderfer M.R.
      • Ooi E.M.M.
      • et al.
      Metabolism and proteomics of large and small dense LDL in combined hyperlipidemia. Effects of rosuvastatin.
      ]. A detailed diagram and explanation of the metabolism of apolipoprotein (apo) B-100 within very low-density lipoproteins, lbLDL, and sdLDL is provided in Supplemental Fig. 3. It should also be mentioned that elevated levels of direct sdLDL-C can be reduced by more than 50% with lifestyle modification and intensive statin therapy [
      • Ai M.
      • Otokozawa S.
      • Asztalos B.F.
      • Nakajima K.
      • Stein E.
      • et al.
      Effects of maximal doses of atorvastatin versus rosuvastatin on small dense low-density lipoprotein cholesterol levels.
      ].

      4.3 Study strengths and limitations

      Strengths of the present study include its prospective nature, the large number of subjects studied, and the inclusion of subjects of both sexes and different ethnic backgrounds (African Americans and non-African Americans, including those of European ancestry, Hispanic/Latino ancestry, and Asian ancestry). The study examined not only ASCVD, but also CHD and ischemic stroke. Limitations of our study were having limited numbers of African Americans and insufficient numbers of Hispanic and Asian subjects to evaluate ASCVD risk in these subgroups. Further studies on these ethnic groups are warranted.

      4.4 Conclusions

      Our data indicate that subjects with direct sdLDL-C top quartile values and those with values ≥ 50 mg/dL, compared to bottom quartile values or those <25 mg/dL, respectively, have about a 50% increased ASCVD and CHD risk in men, women, non-African Americans, and African Americans, after controlling for all standard risk factors. However, this was not the case for calculated sdLDL-C. We believe elevated direct sdLDL-C ≥50 mg/dL should be considered as an important ASCVD risk-enhancer.

      Financial support

      EJS was supported by the U.S. Department of Agriculture – Agricultural Research Service Specific Cooperative Agreements #58-1950-0-014 and #58-1950-4-003 and by grants P50 HL083813-01 and HL117933 from the National Institutes of Health. HI was supported by research grants from the Japan Heart Foundation/Bayer Yakuhin Research Grant Abroad Program, Tokyo, Japan, and from the Denka Corporation, Tokyo, Japan to the Dyslipidemia Foundation of Boston, MA. The statistical consultation and analysis was carried out by EL and C-TL and was supported in part by a grant from the Denka Corporation to the Dyslipidemia Foundation. EL, C-TL, and the Framingham Offspring Study were supported by contracts NHLBI N01-HC 25195 and HHSN268201500001I from the National Institutes of Health. MYT, WG, and the Multi-Ethnic Study of Atherosclerosis were supported by contracts 75N92020D00001, HHSN268201500003I, N01-HC-95159, 75N92020D00005, N01-HC-95160, 75N92020D00002, N01-HC-95161, 75N92020D00003, N01-HC-95162, 75N92020D00006, N01-HC-95163, 75N92020D00004, N01-HC-95164, 75N92020D00007, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-95168 and N01-HC-95169 from the National Institutes of Health, National Heart, Lung, and Blood Institute, and by grants UL1-TR-000040, UL1-TR-001079, and UL1-TR-001420 from the National Center for Advancing Translational Sciences (NCATS). CMB, RCH, and the Atherosclerosis Risk in Communities Study were supported by contract HHSN268201700001I, HHSN268201700002I, HHSN268201700003I, HHSN268201700005I, and HHSN268201700004I from the National Institutes of Health.
      The sponsors had no role in the data analysis or interpretation of this study. The contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of Tufts University, Boston Heart Diagnostics Corporation, Kyushu University, Boston University, Baylor College of Medicine, University of Minnesota School of Public Health, the National Institutes of Health, the U.S. Department of Agriculture Research Service, the National Center for Advancing Translation Sciences, or the Denka Corporation.

      CRediT authorship contribution statement

      Ernst J. Schaefer: Conceptualization, Methodology, Validation, Investigation, Resources, Writing – original draft, Writing – review & editing, Supervision, Project administration, Funding acquisition. Hiroaki Ikezaki: Validation, Investigation, Resources, Writing – review & editing. Margaret R. Diffenderfer: Writing – review & editing, Visualization. Elise Lim: Formal analysis, Data curation, Software. Ching-Ti Liu: Data curation, Software, Formal analysis. Ron C. Hoogeveen: Validation, Investigation, Resources, Data curation, Writing – review & editing. Weihua Guan: Investigation, Resources, Writing – review & editing. Michael Y. Tsai: Conceptualization, Methodology, Validation, Investigation, Resources, Data curation, Writing – review & editing, Supervision, Project administration. Christie M. Ballantyne: Conceptualization, Methodology, Validation, Investigation, Resources, Writing – review & editing, Supervision, Project administration.

      Declaration of competing interest

      EJS has served as a consultant for the Denka Corporation, Tokyo, Japan. RCH has received research grants (to his institution) from Denka Corporation; RCH and CMB are consultants for Denka. None of the other authors has any relevant relationships to disclose.

      Acknowledgements

      We dedicate this manuscript to the memory of our colleague Dr. L. Adrienne Cupples (1945–2022), who helped with the original design of this study. She was best known for her work on the role of genetics in ASCVD risk and served as a professor of biostatistics and epidemiology at Boston University School of Public Health. We also thank the other investigators, the staff, and the participants of the Framingham Offspring Study, the Atherosclerosis Risk in Communities Study, and the Multi-Ethnic Study of Atherosclerosis for their valuable contributions. We also thank Drs. Yasuki Ito and Asako Machida of the Denka Corporation, Tokyo, Japan for their support of this study. The study was presented in part at the annual scientific sessions of the American Heart Association, November 13, 2021, Boston, MA USA.

      Appendix A. Supplementary data

      The following is the Supplementary data to this article:

      References

        • Goff Jr., D.C.
        • Lloyd-Jones D.M.
        • Bennett G.
        • Coady S.
        • D'Agostino R.B.
        • et al.
        American College of cardiology/American heart association task force on practice guidelines. 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of cardiology/American heart association task force on practice guidelines.
        Circulation. 2014; 129: S49-S73https://doi.org/10.1161/01.cir.0000437741.48606.98
        • Grundy S.M.
        • Stone N.J.
        • Bailey A.L.
        • Beam C.
        • Birtcher K.K.
        • et al.
        AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/ADA/APHA/ASPC/NLA/PCNA guideline on the management of blood cholesterol: a report of the American College of cardiology/American heart association task force on clinical practice guidelines.
        Circulation. 2018; 139 (2019): e1082-e1143https://doi.org/10.1161/CIR.0000000000000625
        • Silverman M.G.
        • Ference B.A.
        • Im K.
        • Wiviott S.D.
        • Giugliano R.P.
        • et al.
        Association between lowering LDL-C and cardiovascular risk reduction among different therapeutic interventions: a systematic review and meta-analysis.
        JAMA. 2016; 316: 1289-1297https://doi.org/10.1001/jama.2016.13985
        • Schaefer E.J.
        • Tsunoda F.
        • Diffenderfer M.R.
        • Polisecki E.A.
        • Thai N.
        • et al.
        The measurement of lipids, lipoproteins, apolipoproteins, fatty acids, and sterols, and next generation sequencing for the diagnosis and treatment of lipid disorders.
        in: Feingold K.R. Anawalt B. Boyce A. Chrousos G. Dungan K. Endotext [Internet]. MDText.com, Inc, South Dartmouth, MA2016: 1-69
        • Austin M.A.
        • King M.C.
        • Vranizan K.M.
        • Krauss R.M.
        Atherogenic lipoprotein phenotype. A proposed genetic marker for coronary heart disease risk.
        Circulation. 1990; 82: 495-506https://doi.org/10.1161/01.cir.82.2.495
        • Campos H.
        • Genest J.J.
        • Blijlevens E.
        • McNamara J.R.
        • Jenner J.
        • et al.
        Low-density lipoprotein particle size and coronary artery disease.
        Arterioscler. Thromb. 1992; 12: 187-195https://doi.org/10.1161/01.atv.12.2.187
        • Mora S.
        • Szklo M.
        • Otvos J.D.
        • Greenland P.
        • Psaty B.M.
        • et al.
        LDL particle subclasses, LDL particle size, and carotid atherosclerosis in the Multi-Ethnic Study of Atherosclerosis (MESA).
        Atherosclerosis. 2007; 192: 211-217https://doi.org/10.1016/j.atherosclerosis.2006.05.007
        • Mora S.
        • Otvos J.D.
        • Rifai N.
        • Rosenson R.S.
        • Buring J.E.
        • et al.
        Lipoprotein particle profiles by nuclear magnetic resonance compared with standard lipids and apolipoproteins in predicting incident cardiovascular disease in women.
        Circulation. 2009; 119: 931-939https://doi.org/10.1161/CIRCULATIONAHA.108.816181
        • Mora S.
        • Caulfield M.P.
        • Wohlgemuth J.
        • Chen Z.
        • Superko H.R.
        • et al.
        Atherogenic lipoprotein subfractions determined by ion mobility and first cardiovascular events after random allocation to high-intensity statin or placebo: the JUPITER Trial.
        Circulation. 2015; 132: 2220-2229https://doi.org/10.1161/CIRCULATIONAHA.115.016857
        • Burstein M.
        • Scholnick H.R.
        • Morfin R.
        Rapid method for the isolation of lipoproteins from human serum by precipitation with polyanions.
        J. Lipid Res. 1970; 11: 583-595
      1. Manual of Laboratory Operations. Lipid Research Clinics Program. Lipid and Lipoprotein Analysis. DHEW Pub (NIH), 1975: 75-628
        • McNamara J.R.
        • Schaefer E.J.
        Automated enzymatic standardized lipid analyses for plasma and lipoprotein fractions.
        Clin. Chim. Acta. 1987; 166: 1-8https://doi.org/10.1016/0009-8981(87)90188-4
        • Sugiuchi H.
        • Uji Y.
        • Okabe H.
        • Irie T.
        • Uekama K.
        • et al.
        Direct measurement of high-density lipoprotein cholesterol in serum with polyethylene glycol-modified enzymes and sulfated α-cyclodextrin.
        Clin. Chem. 1995; 41: 717-723
        • Friedewald W.T.
        • Levy R.I.
        • Fredrickson D.S.
        Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge.
        Clin. Chem. 1972; 18: 499-502
        • Martin S.S.
        • Blaha M.J.
        • Elshazly M.B.
        • Toth P.P.
        • Kwiterovich P.O.
        • et al.
        Comparison of a novel method vs the Friedewald equation for estimating low-density lipoprotein cholesterol levels from the standard lipid profile.
        JAMA. 2013; 310: 2061-2068https://doi.org/10.1001/jama.2013.280532
        • Meeusen J.W.
        • Lueke A.J.
        • Jaffe A.S.
        • Saenger A.K.
        Validation of a proposed novel equation for estimating LDL cholesterol.
        Clin. Chem. 2014; 60: 1519-1523https://doi.org/10.1373/clinchem.2014.227710
        • Sampson M.
        • Ling C.
        • Sun Q.
        • Harb R.
        • Ashmaig M.
        • et al.
        A new equation for calculation of low-density lipoprotein cholesterol in patients with normolipemia and/or hypertriglyceridemia.
        JAMA Cardiol. 2021; 5: 540-548https://doi.org/10.1001/jamacardio.2020.0013
        • McNamara J.R.
        • Cole T.G.
        • Contois J.H.
        • Ferguson C.A.
        • Ordovas J.M.
        • et al.
        Immunoseparation method for measuring low-density lipoprotein cholesterol directly from serum evaluated.
        Clin. Chem. 1995; 41: 232-240
        • Okada M.
        • Matsui H.
        • Ito Y.
        • Fujiwara A.
        • Inano K.
        Low density lipoprotein cholesterol can be chemically measured; a new superior method.
        J. Lab. Clin. Med. 1998; 132: 195-201https://doi.org/10.1016/s0022-2143(98)90168-8
        • Sakaue T.
        • Hirano T.
        • Yoshino G.
        • Sakai K.
        • Takeuchi H.
        • et al.
        Reactions of direct LDL-cholesterol assays with pure LDL fraction and IDL: comparison of three homogeneous methods.
        Clin. Chim. Acta. 2000; 295: 97-106https://doi.org/10.1016/s0009-8981(00)00200-x
        • Otokozawa S.
        • Ai M.
        • Asztalos B.F.
        • White C.C.
        • Demissie-Banjaw S.
        • et al.
        Direct assessment of plasma low-density lipoprotein and high-density lipoprotein cholesterol and coronary heart disease: results from the Framingham Offspring Study.
        Atherosclerosis. 2010; 213: 251-255https://doi.org/10.1016/j.atherosclerosis.2010.02.041
        • Hirano T.
        • Ito Y.
        • Saegusa H.
        • Yoshino G.
        A novel and simple method for quantification of small dense LDL.
        J. Lipid Res. 2003; 44: 2193-2201https://doi.org/10.1194/jlr.D300007-JLR200
        • Ito Y.
        • Fujimura M.
        • Ohta M.
        • Hirano T.
        Development of a homogeneous assay for measurement of small dense LDL cholesterol.
        Clin. Chem. 2011; 57: 57-65https://doi.org/10.1373/clinchem.2010.149559
        • Hirano T.
        • Ito Y.
        • Koba S.
        • Toyoda Ikejiri A.
        • et al.
        Clinical significance of small dense low-density lipoprotein cholesterol levels determined by the simple precipitation method.
        Arterioscler. Thromb. Vasc. Biol. 2004; 24: 558-563https://doi.org/10.1161/01.ATV.0000117179.92263.08
        • Koba S.
        • Yokota Y.
        • Hirano T.
        • Ito Y.
        • Ban Y.
        • et al.
        Small LDL-cholesterol is superior to LDL-cholesterol for determining severe coronary atherosclerosis.
        J. Atherosclerosis Thromb. 2008; 15: 350-360https://doi.org/10.5551/jat.e572
        • Ai M.
        • Otokozawa S.
        • Asztalos B.F.
        • Ito Y.
        • Nakajima K.
        • et al.
        Small dense LDL cholesterol and coronary heart disease: results from the Framingham Offspring Study.
        Clin. Chem. 2010; 56: 967-976https://doi.org/10.1373/clinchem.2009.137489
        • Nishikura T.
        • Koba S.
        • Yokota Y.
        • Hirano T.
        • Tsunoda F.
        • et al.
        Elevated small dense low-density lipoprotein cholesterol as a predictor for future cardiovascular events in patients with stable coronary artery disease.
        J. Atherosclerosis Thromb. 2014; 21: 755-767https://doi.org/10.5551/jat.23465
        • Arai H.
        • Kokubo Y.
        • Watanabe M.
        • Sawamura T.
        • Ito Y.
        • et al.
        Small dense low-density lipoproteins cholesterol can predict incident cardiovascular disease in an urban Japanese cohort: the Suita study.
        J. Atherosclerosis Thromb. 2013; 20: 195-203https://doi.org/10.5551/jat.14936
        • Tsai M.Y.
        • Steffen B.T.
        • Guan W.
        • McClelland R.L.
        • Warnick R.
        • et al.
        New automated assay of small dense low-density lipoprotein cholesterol identifies risk of coronary heart disease: the Multi-ethnic Study of Atherosclerosis.
        Arterioscler. Thromb. Vasc. Biol. 2014; 34: 196-201https://doi.org/10.1161/ATVBAHA.113.302401
        • Hoogeveen R.C.
        • Gaubatz J.W.
        • Sun W.
        • Dodge R.C.
        • Crosby J.R.
        • et al.
        Small dense low-density lipoprotein-cholesterol concentrations predict risk for coronary heart disease: the Atherosclerosis Risk in Communities (ARIC) study.
        Arterioscler. Thromb. Vasc. Biol. 2014; 34: 1069-1077https://doi.org/10.1161/ATVBAHA.114.303284
        • Balling M.
        • Nordestgaard B.G.
        • Langsted A.
        • Varbo A.
        • Kamstrup P.R.
        • et al.
        Small dense low-density lipoprotein cholesterol predicts atherosclerotic cardiovascular disease in the Copenhagen General Population Study.
        J. Am. Coll. Cardiol. 2020; 75: 2873-2875https://doi.org/10.1016/j.jacc.2020.03.072
        • Ikezaki H.
        • Lim E.
        • Cupples L.A.
        • Liu C.T.
        • Asztalos B.F.
        • et al.
        Small dense low‐density lipoprotein cholesterol is the most atherogenic lipoprotein parameter in the prospective Framingham Offspring Study.
        J. Am. Heart Assoc. 2021; 10e019140https://doi.org/10.1161/JAHA.120.019140
        • Sampson M.
        • Wolska A.
        • Warnick R.
        • Lucero D.
        • Remaley A.T.
        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.
        Clin. Chem. 2021; 67: 987-997https://doi.org/10.1093/clinchem/hvab048
        • Pencina M.J.
        • D'Agostino R.B.
        Evaluating discrimination of risk prediction models: the C Statistic.
        JAMA. 2015; 314 (Sr): 1063-1064https://doi.org/10.1001/jama.2015.11082
        • Ikezaki H.
        • Fisher V.A.
        • Lim E.
        • Ai M.
        • Liu C.T.
        • et al.
        Direct versus calculated low-density lipoprotein cholesterol and C reactive protein in cardiovascular disease risk assessment in the Framingham Offspring Study.
        Clin. Chem. 2019; 65: 1102-1114https://doi.org/10.1373/clinchem.2019.304600
        • Thongtang N.
        • Diffenderfer M.R.
        • Ooi E.M.M.
        • et al.
        Metabolism and proteomics of large and small dense LDL in combined hyperlipidemia. Effects of rosuvastatin.
        J. Lipid Res. 2017; 58: 1315-1324https://doi.org/10.1194/jlr.M073882
        • Ai M.
        • Otokozawa S.
        • Asztalos B.F.
        • Nakajima K.
        • Stein E.
        • et al.
        Effects of maximal doses of atorvastatin versus rosuvastatin on small dense low-density lipoprotein cholesterol levels.
        Am. J. Cardiol. 2008; 101: 315-318https://doi.org/10.1016/j.amjcard.2007.08.035