SUPPLEMENTAL MATERIAL

Similar documents
Ischemic Heart and Cerebrovascular Disease. Harold E. Lebovitz, MD, FACE Kathmandu November 2010

University of Bristol - Explore Bristol Research

Optimizing risk assessment of total cardiovascular risk What are the tools? Lars Rydén Professor Karolinska Institutet Stockholm, Sweden

CVD Prevention, Who to Consider

There are many ways to lower triglycerides in humans: Which are the most relevant for pancreatitis and for CV risk?

Supplementary Online Content

Review of guidelines for management of dyslipidemia in diabetic patients

Case Presentation. Rafael Bitzur The Bert W Strassburger Lipid Center Sheba Medical Center Tel Hashomer

Dyslipidemia in women: Who should be treated and how?

HIGH LDL CHOLESTEROL IS NOT AN INDEPENDENT RISK FACTOR FOR HEART ATTACKS AND STROKES

CVD risk assessment using risk scores in primary and secondary prevention

Fasting or non fasting?

Lipid Panel Management Refresher Course for the Family Physician

9/29/2015. Primary Prevention of Heart Disease: Objectives. Objectives. What works? What doesn t?

The Framingham Coronary Heart Disease Risk Score

EUROPEAN SURVEY OF CARDIOVASCULAR DISEASE PREVENTION AND DIABETES EUROASPIRE IV. GUY DE BACKER Ghent University,Belgium

High-Density Lipoprotein Subclass Testing in the Diagnosis and Management of Cardiovascular Disease. Original Policy Date

Behind LDL: The Metabolism of ApoB, the Essential Apolipoprotein in LDL and VLDL

John J.P. Kastelein MD PhD Professor of Medicine Dept. of Vascular Medicine Academic Medial Center / University of Amsterdam

Dyslipidemia: Lots of Good Evidence, Less Good Interpretation.

Proven and Proposed Cardiovascular Benefits of Soyfoods

Kidney and heart: dangerous liaisons. Luis M. RUILOPE (Madrid, Spain)

Cardiovascular Complications of Diabetes

Traditional Asian Soyfoods. Proven and Proposed Cardiovascular Benefits of Soyfoods. Reduction (%) in CHD Mortality in Eastern Finland ( )

Antihypertensive Trial Design ALLHAT

Assessing Cardiovascular Risk to Optimally Stratify Low- and Moderate- Risk Patients. Copyright. Not for Sale or Commercial Distribution

The Clinical Unmet need in the patient with Diabetes and ACS

10/17/16. Assessing cardiovascular risk through use of inflammation testing

Lipoprotein (a) Disclosures 2/20/2013. Lipoprotein (a): Should We Measure? Should We Treat? Health Diagnostic Laboratory, Inc. No other disclosures

The Whitehall II study originally comprised 10,308 (3413 women) individuals who, at

Advanced IPD meta-analysis methods for observational studies

9/18/2017 DISCLOSURES. Consultant: RubiconMD. Research: Amgen, NHLBI OUTLINE OBJECTIVES. Review current CV risk assessment tools.

The JUPITER trial: What does it tell us? Alice Y.Y. Cheng, MD, FRCPC January 24, 2009

The inhibition of CETP: From simply raising HDL-c to promoting cholesterol efflux and lowering of atherogenic lipoproteins Prof Dr J Wouter Jukema

An example of a systematic review and meta-analysis

Total risk management of Cardiovascular diseases Nobuhiro Yamada

Update on Dyslipidemia and Recent Data on Treating the Statin Intolerant Patient

Environmental. Vascular / Tissue. Metabolics

2/17/2015. ApoB Versus Non-HDL Cholesterol And the Winner Is. Braunwald10 th Edition Chapter 45 J GenestPA Libby. Emerging Risk Factor Collaboration

Apolipoprotein B in the Risk Assessment and Management of Cardiovascular Disease. Original Policy Date

LIPOPROTEIN PROFILING

Should we prescribe aspirin and statins to all subjects over 65? (Or even all over 55?) Terje R.Pedersen Oslo University Hospital Oslo, Norway

Estrogens vs Testosterone for cardiovascular health and longevity

Preventing Myocardial Infarction in the Young Adult in the First Place: How Do the National Cholesterol Education Panel III Guidelines Perform?

Low-density lipoproteins cause atherosclerotic cardiovascular disease (ASCVD) 1. Evidence from genetic, epidemiologic and clinical studies

Appendix This appendix was part of the submitted manuscript and has been peer reviewed. It is posted as supplied by the authors.

2013 Cholesterol Guidelines. Anna Broz MSN, RN, CNP, AACC Adult Certified Nurse Practitioner North Ohio Heart, Inc.

Inflammation and and Heart Heart Disease in Women Inflammation and Heart Disease

The target blood pressure in patients with diabetes is <130 mm Hg

Guidelines on cardiovascular risk assessment and management

Dyslipidemia in the light of Current Guidelines - Do we change our Practice?

What have We Learned in Dyslipidemia Management Since the Publication of the 2013 ACC/AHA Guideline?

Supplementary Online Content

Prevention of Cardiovasular Diseases

Appendix: Definition of variables in the studies included in the meta-analysis

LDL cholesterol and cardiovascular outcomes?

Placebo-Controlled Statin Trials MANAGEMENT OF HIGH BLOOD CHOLESTEROL MANAGEMENT OF HIGH BLOOD CHOLESTEROL: IMPLICATIONS OF THE NEW GUIDELINES

ALLHAT Role of Diuretics in the Prevention of Heart Failure - The Antihypertensive and Lipid- Lowering Treatment to Prevent Heart Attack Trial

Biomarkers (Novel) in Risk Assessment and Management of Cardiovascular Disease

The Diabetes Link to Heart Disease

Biases in clinical research. Seungho Ryu, MD, PhD Kanguk Samsung Hospital, Sungkyunkwan University

Low-density lipoprotein as the key factor in atherogenesis too high, too long, or both

Preventive Cardiology Scientific evidence

Which CVS risk reduction strategy fits better to carotid US findings?

Subclinical atherosclerosis in CVD: Risk stratification & management Raul Santos, MD

김광일 서울대학교의과대학내과학교실 분당서울대학교병원내과

1. Which one of the following patients does not need to be screened for hyperlipidemia:

ARIC Manuscript Proposal # PC Reviewed: 2/10/09 Status: A Priority: 2 SC Reviewed: Status: Priority:

I t is well established that non-insulin dependent diabetes is

Novel Biomarkers in Risk Assessment and Management of Cardiovascular Disease

well-targeted primary prevention of cardiovascular disease: an underused high-value intervention?

Welcome! Mark May 14, Sat!

Andrew Cohen, MD and Neil S. Skolnik, MD INTRODUCTION

Lipid Risk Factors (Novel) in Risk Assessment and Management of Cardiovascular Disease

Cedars Sinai Diabetes. Michael A. Weber

Epidemiological studies indicate that a parental or family

Update on CVD and Microvascular Complications in T2D

The apob/apoa-i ratio: a strong, new risk factor for. cardiovascular disease and a target for lipidlowering

Placebo-Controlled Statin Trials EXPLAINING THE DECREASE IN DEATHS FROM CHD! PREVENTION OF CARDIOVASCULAR DISEASE IN WOMEN EXPLAINING THE DECREASE IN

Population models of health impact of combination polypharmacy

Supplementary Online Content

Downloaded from:

Placebo-Controlled Statin Trials Prevention Of CVD in Women"

Placebo-Controlled Statin Trials

ARE YOU AT RISK OF A HEART ATTACK OR STROKE? Understand How Controlling Your Cholesterol Reduces Your Risk

Supplementary Online Content

NATIONAL INSTITUTE FOR HEALTH AND CLINICAL EXCELLENCE

Statins in the elderly : Is there a rationale?

Disclosures. Background 1 What is Known MENOPAUSE, ESTROGENS, AND LIPOPROTEIN PARTICLES. Background 2 What is Not Known 10/2/2017

Treatment of Cardiovascular Risk Factors. Kevin M Hayes D.O. F.A.C.C. First Coast Heart and Vascular Center

1. Albuminuria an early sign of glomerular damage and renal disease. albuminuria

The CARI Guidelines Caring for Australians with Renal Impairment. Cardiovascular Risk Factors

CRP for the Clinician

American Journal of Internal Medicine

THE CRITICAL ROLE OF NURSES. Helping patients take control of their LDL-C to lower the risk of MI and stroke.

How to Reduce Residual Risk in Primary Prevention

Kathryn M. Rexrode, MD, MPH. Assistant Professor. Division of Preventive Medicine Brigham and Women s s Hospital Harvard Medical School

Soo LIM, MD, PHD Internal Medicine Seoul National University Bundang Hospital

HDL-C. J Jpn Coll Angiol, 2008, 48: NIPPON DATA80, MEGA study, JELIS, dyslipidemia, risk assessment chart

Transcription:

SUPPLEMENTAL MATERIAL A Meta-analysis of LDL-C, non-hdl-c, and apob as markers of cardiovascular risk. Slide # Contents 2 Table A1. List of candidate reports 8 Table A2. List of covariates/model adjustments 9 Figure A1. Forest plot of LDL-C relative risk ratios 10 Figure A2. Forest plot of non-hdl-c relative risk ratios 11 Figure A3. Forest plot of relative apob risk ratios 12 Figure A4. Log apob RRR log non-hdl-c RRR forest plot 13 Figure A5. Log non-hdl-c RRR log LDL-C RRR forest plot 14 Figure A6. Log apob RRR log LDL-C RRR forest plot 15 Figure A7. Subgroup analyses by study attributes 16 Figure A8. Selection bias assessments: funnel plots of precision by difference in log relative risk ratios Note: Citations throughout this appendix refer to the reference list of the main article. 1

Table A1. List of candidate reports by inclusion/exclusion categories #A: ERFC-identified report without the requisite non-hdl-c and apob data B: ERFC-identified study with requisite data but without published non-hdl-c and apob relative risk ratios (RRRs) C: ERFC-identified report with published RRRs not included in ERFC meta-analysis D: Reports from ERFC-cited studies included in both analyses E: Reports from studies not cited by ERFC with RRRs F: Reports not cited by ERFC without RRRs 2 Ref # meta-analyses Study citation 1 A Neither *Air Force/Texas Coronary Atherosclerosis Prevention Study Circulation 2000; 101(5):477-484. 2 A Neither *Antihypertensive and Lipid-Lowering to prevent Heart Attacks JAMA 2002; 288(23):2998-3007. 3 A Neither *British Regional Heart Study BMJ 1989; 298(6679):998-1002. 4 A Neither *British Women's Heart and Health Study J Epidemiol Community Health 2003; 57(2):134-140. 5 A Neither *Busselton population health J Epidemiol Community Health 1997; 51(5):515-519. 6 A Neither *Caerphilly and Speedwell Collaborative Heart Disease Br Heart J 1992; 68(1):60-66. 7 A Neither *Cardiovascular Health Study J Am Geriatr Soc 2004; 52(10):1639-1647. 8 A Neither *Edinburgh Artery Study Am J Epidemiol 1992; 135(4):331-340. 9 A Neither *Epidemiologische Studie zu Chancen der Verhütung, Pharmacoepidemiol Drug Saf 2008; 17(5):485-494. 10 A Neither *Finish Risk Cohort Study Int J Epidemiol 2000; 29(1):49-56. 11 A Neither *Finland, Italy and Netherlands Elderly Study Eur Heart J 2001;22(7):573-579. 12 A Neither *GOTMONICA J Intern Med 1997; 242(3):199-211. 13 A Neither *Glostrup Study Int J Epidemiol 1991; 20(1):105-113. 14 A Neither *Göteborg 1943 Study J Intern Med 2000; 247(1):111-118. 15 A Neither *Honolulu Heart Program Am J Cardiol 2000; 86(4):412-416. 16 A Neither *Hoorn Study Diabetologia 2003; 46(7):910-916. 17 A Neither *Israeli IHD/Glucose Intolerance, Obesity, Hypertension Preventive Medicine 2005; 41(1):85-91. 18 A Neither *Japanese Am J Epidemiol 2001; 153(5):490-499. 19 A Neither *Lower Extremity Arterial Disease Event Reduction Trial Curr Control Trials Cardiovasc Med 2001; 2(4):195 20 A Neither *Multiple Risk Factor Intervention Trial 1 Am J Cardiol 1986; 57(8):538-545. 21 A Neither *Northwick Park Heart Study II Atherosclerosis 2005; 181(1):93-100. *Report cited by ERFC as providing data on triglycerides, HDL-C and non-hdl-c, and other risk factors

Table A1. List of candidate reports by inclusion/exclusion categories #A: ERFC-identified report without the requisite non-hdl-c and apob data B: ERFC-identified study with requisite data but without published non-hdl-c and apob relative risk ratios (RRRs) C: ERFC-identified report with published RRRs not included in ERFC meta-analysis D: Reports from ERFC-cited studies included in both analyses E: Reports from studies not cited by ERFC with RRRs F: Reports not cited by ERFC without RRRs 3 Ref # meta-analyses Study citation 22 A Neither *Population Study of Women in Göteborg, Sweden Circulation 2004; 109(5):601-606. 23 A Neither *Progetto CUORE Eur J Cardiovasc Prev Rehabil 2006; 13(4):562-570 24 A Neither *Rancho Bernardo Study Am J Cardiol 2003; 91(11):1311-1315. 25 A Neither *Reykjavik Study J Cardiovasc Risk 2002; 9(2):67-76. 26 A Neither *Risk Factors and Life Expectancy Pooling Project Eur J Epidemiol 1993; 9(5):459-476. 27 A Neither *Scottish Heart Health Extended Cohort Heart 2007; 93(2):172-176. 28 A Neither *Established Populations for Epidemiologic Study of Elderly Aging (Milano ) 1993; 5(1):27-37. 29 A Neither *Tromsø Study Lancet 1977; 1(8019):965-968. 30 A Neither *West of Scotland Coronary Prevention Study Circulation 1998; 97(15):1440-1445. 31 A Neither *Yao city Japan J Clin Epidemiol 1994; 47(9):961-969. 32 A Neither *Zaragosa study BMC Public Health 2006; 6:38. 33 A Neither *Zutphen Elderly Study Am J Epidemiol 1996; 143(2):151-158. 34 B ERFC *Atherosclerosis Risk in Communities Study Circulation 104(10):1108-1113. 35 B ERFC *BUPA Study Lancet. 1994;343:75 79 36 B ERFC *Bruneck Study Arterioscler Thromb Vasc Biol 1999; 19(6):1484-14 37 B ERFC *Cardiovascular Study in the Elderly J Hum Hypertens 1998; 12(9):575-581. 38 B ERFC *Diet and Risk of Cardiovascular Disease in Spain Annals of Nutrition and Metabolism 2000; 44(3):10 39 B ERFC *Dubbo Study of the Elderly Am J Cardiol 2002; 89(1):69-72. 40 B ERFC *EPIC Norfolk Study Ann Intern Med 2007; 146(9):640-648. 41 B ERFC *Göttingen Risk Incidence and Prevalence Study Atherosclerosis 1997; 129(2):221-230. 42 B ERFC *Kuopio Ischaemic Heart Disease Study Circulation 1991; 84(1):129-139. *Report cited by ERFC as providing data on triglycerides, HDL-C and non-hdl-c, and other risk factors

Table A1. List of candidate reports by inclusion/exclusion categories #A: ERFC-identified report without the requisite non-hdl-c and apob data B: ERFC-identified study with requisite data but without published non-hdl-c and apob relative risk ratios (RRRs) C: ERFC-identified report with published RRRs not included in ERFC meta-analysis D: Reports from ERFC-cited studies included in both analyses E: Reports from studies not cited by ERFC with RRRs F: Reports not cited by ERFC without RRRs 4 Ref # meta-analyses Study citation 43 B ERFC *National Health and Nutrition Examination Survey III Am J Cardiol 2000; 86(3):299-304. 44 B ERFC *Prospective Cardiovascular Münster Study Am J Cardiol 1992; 70(7):733-737. 45 B ERFC *Prospective Epidemiological Study of Myocardial Infarction Arterioscler Thromb Vasc Biol 2002; 22(7):1155-11 46 B ERFC *Prospective Study of Pravastatin in the Elderly at Risk Circulation 2005; 112(20):3058-3065. 47 B ERFC *Quebec Cardiovascular Study Atherosclerosis 2000; 153(2):263-272. 48 B ERFC *Strong Heart Study Diabetes 1992; 41 Suppl 2:4-11.:4-11. 49 B ERFC *Turkish Adult Risk Factor Study J Epidemiol Community Health 1992; 46(5):470-476. 50 B ERFC *Uppsala Arterioscler Thromb Vasc Biol 2006; 26(2):406-410 51 B ERFC *VIP/MONICA J Intern Med 1998; 244(5):425-430. 52 B ERFC *Whitehall II Study Arterioscler Thromb Vasc Biol 2008; 28(8):1556-15 53 C Current *Framingham Offspring Study JAMA. 2007;298:776 785 54 D ERFC *Copenhagen City Heart Study JAMA 2007; 298(3):299-308. 55 D Current Copenhagen City Heart Study Arterioscler Thromb Vasc Biol. 2007;27:661 670 56 D Both *Nurses Health Study (11) Circulation.2004;110:2824 2830 57 D ERFC *Women s Health Study (15) JAMA. 2005;294:326 3633 58 D Current Women s Health Study (12) Circulation. 2009;119:931-939. 59 E Current Apolipoprotein Related Mortality Risk Study (21) J Intern Med. 2008;264:30 38 60 E Current Casale Monferrato Study (19) Diabetologia. 2006;49:937 944 61 E Current Chin-Shan Cohort (20) J Lipid Res. 2007;48:2499 2505 62 E Current Health Professionals Follow-up Study (16) Diabetes Care. 2004;27: 1991 1997 63 E Current Health Professionals Follow-up Study (18) Circulation. 2005;1112:3375 3583 *Report cited by ERFC as providing data on triglycerides, HDL-C and non-hdl-c, and other risk factors

Table A1. List of candidate reports by inclusion/exclusion categories #A: ERFC-identified report without the requisite non-hdl-c and apob data B: ERFC-identified study with requisite data but without published non-hdl-c and apob relative risk ratios (RRRs) C: ERFC-identified report with published RRRs not included in ERFC meta-analysis D: Reports from ERFC-cited studies included in both analyses E: Reports from studies not cited by ERFC with RRRs F: Reports not cited by ERFC without RRRs 5 Ref # meta-analyses Study citation 64 E Current INTERHEART Study (22) Lancet 2008; 372: 224 33 65 E Current International Studies of Infarct Survival (23) Euro Heart J (2009) 30, 2137 2146 66 E Current Monitoring of Trends and Determinants in CVD (17) EurHeart J. 2005;26:271 278 67 F Neither 4S Placebo Wing clinical trial Circulation. 1998;97:1453 1460 68 F Neither Air Force/Texas Coronary Atherosclerosis Prevention Study Circulation. 2000;101:477 484 69 F Neither Angina Prognosis Study in Stockholm Atherosclerosis. 1997;135:109 118 70 F Neither ApoB and Type 1 Diabetes J Immunol Methods. 2006;260: 272 280 71 F Neither Apolipoprotein Related Mortality Risk Study Clin Chem Lab Med 2004; 42: 1355 63. 72 F Neither Apolipoprotein Related Mortality Risk Study Lancet. 2001;358:2026 2033 73 F Neither Apolipoprotein Related Mortality Risk Study Lancet. 2003;361: 777 780 74 F Neither Bezafibrate Coronary Atherosclerosis Intervention Trial J Am Coll Cardiol. 1998;32:1648 1656 75 F Neither Bugalusu Heart Study Pediatrics. 2008;121:924 929 76 F Neither Caerphilly Prospective Study Eur J Clin Invest 2000; 30: 947 56. 77 F Neither Cardiovascular Health Study Arterioscler Thromb Vasc Biol.2002;22:1175 1180 78 F Neither Cardiovascular risk in Young Finns Study J Am CollCardiol. 2008;1:1 2 79 F Neither Diabetes Atherosclerosis Intervention Study Circulation. 2003;107:1733 1737 80 F Neither Dubbo Study of the Elderly Atherosclerosis 2001; 159: 201 8. 81 F Neither EPIC Norfolk Study J Am Coll Cardiol.2007;49:547 553 82 F Neither Eastern Finland Heart Survey Am J Cardiol. 1985;56:228 231 83 F Neither Fenofibrate Intervention and Event Lowering in Diabetes Diabetologia 2010 53:1846-1855 84 F Neither Framingham Offspring Study J Clin Lipidol. 2007;1:583 592 *Report cited by ERFC as providing data on triglycerides, HDL-C and non-hdl-c, and other risk factors

Table A1. List of candidate reports by inclusion/exclusion categories #A: ERFC-identified report without the requisite non-hdl-c and apob data B: ERFC-identified study with requisite data but without published non-hdl-c and apob relative risk ratios (RRRs) C: ERFC-identified report with published RRRs not included in ERFC meta-analysis D: Reports from ERFC-cited studies included in both analyses E: Reports from studies not cited by ERFC with RRRs F: Reports not cited by ERFC without RRRs 6 Ref # meta-analyses Study citation 85 F Neither Glostrup Study Atherosclerosis 1997; 132: 77 84. 86 F Neither Guernsey Atherosclerosis 1992; 92: 177 85. 87 F Neither Incremental Decrease in End points Through Aggressive Lipidlowering Ann Med. 2008:1 9 88 F Neither Kuopio Ischaemic Heart Disease Study Circulation 1992; 86: 803 11. 89 F Neither Kuopio Ischaemic Heart Disease Study Arterioscler Thromb Vasc Biol 1999; 19: 2742 8. 90 F Neither Leiden Heart Study. Arterioscler Thromb Vasc Biol. 2000;20: 2408 2413 91 F Neither Long-term Intervention with Pravastatin in Ischaemic Disease Circulation. 2002;105: 1162 1169 92 F Neither Monitoring of Trends and Determinants in CVD Eur Heart J 2005;26: 271 8. 93 F Neither National Health and Nutrition Examination Survey Am J Cardiol. 2006;98:1047 1052 94 F Neither Northwick Park Heart Study Arterioscler Thromb Vasc Biol.2002;22:1918 1923 95 F Neither Quebec Cardiovascular Study Arterioscler Thromb Vasc Biol 2005; 25:553 9. 96 F Neither Quebec Cardiovascular Study Circulation. 1996;94:273 278 97 F Neither Quebec Cardiovascular Study Am J Cardiol. 2006;97:997 1001 98 F Neither Reykjavik Study Am J Cardiol 1992; 69: 1251 4. 99 F Neither Simon et al Atherosclerosis.2005;179:339 344 100 F Neither TEKHARF Survey Anadolu Kardiyol Derg. 2007;2:128 133 101 F Neither THROMBO Metabolic Syndrome Atherosclerosis. 2004;177:367 373 102 F Neither TNT/IDEAL Circulation. 2008; 117:3002 3009 103 F Neither Thrombosis Prevention Trial Circulation. 1999;99:2517 2522 104 F Neither U.S. Physicians Health Study N Engl J Med 1991; 325: 373 81. *Report cited by ERFC as providing data on triglycerides, HDL-C and non-hdl-c, and other risk factors

Table A1. List of candidate reports by inclusion/exclusion categories #A: ERFC-identified report without the requisite non-hdl-c and apob data B: ERFC-identified study with requisite data but without published non-hdl-c and apob relative risk ratios (RRRs) C: ERFC-identified report with published RRRs not included in ERFC meta-analysis D: Reports from ERFC-cited studies included in both analyses E: Reports from studies not cited by ERFC with RRRs F: Reports not cited by ERFC without RRRs 7 Ref # meta-analyses Study citation 105 F Neither Uppsala Longitudinal Study of Adult Men Am Heart J 2004; 148: 596 601. 106 F Neither VA_HIT Circulation. 2006;113: 1556 1563 107 F Neither Women s Health Study Circulation.2002;106:1930 1937 *Report cited by ERFC as providing data on triglycerides, HDL-C and non-hdl-c, and other risk factors

Table A2. List of covariates/model adjustments Study abbreviation Health Professionals (16) Nurses' Health(11) MONICA/KORA(17) Health Professionals(18) Women's Health(15) Casale Monferrato(19) Copenhagen Heart(13) Chin-Shan Cohort(20) Fram Offspring (10) AMORIS(21) INTERHEART(22) Women's Health(12) ISIS(23) ADJUSTMENTS age, BMI, family history of MI, physical activity, smoking (never, past, or current), alcohol consumption, fasting status, history of hypertension, aspirin use, HbA1c age, smoking, fasting status, CRP, homocysteine, BMI, family history, hypertension, diabetes, hormone use, physical activity, alcohol intake, blood draw parameters age, diabetes, regular smoking, BMI, alcohol intake age, smoking, blood draw, BMI, parent MI < 60, hypertension, alcohol intake, physical activity age, Framingham BP category, BMI, diabetes, current smoking age, sex, hypertension, smoking, CHD, AER, fibrinogen, cumulative average HbA1c, referring physician age age group, gender, BMI, smoking, alcohol, marital status, education, occupation, exercise, family CHD history, hypertension, diabetes age, SBP, antihypertensive treatment, diabetes, smoking age sex age, sex, smoking, geographic region age, randomized treatment assignment, smoking status, postmenopausal hormone use, blood pressure, diabetes, BMI age, sex, smoking, BMI 8

LDL C relative risk ratio (RRR) Study name Subgroup within study Statistics for each study Hazard ratio and 95% CI Hazard Lower Upper ratio limit limit Z-Value p-value Health Professionals (16) men 1.191 0.979 1.448 1.749 0.080 Nurses' Health(11) women 1.400 1.212 1.617 4.585 0.000 Health Professionals (18) men 1.297 1.080 1.557 2.787 0.005 Casale Monferrato(19) pooled 0.902 0.779 1.045-1.369 0.171 Copenhagen Heart(13) men 1.276 1.114 1.461 3.517 0.000 Copenhagen Heart(13) women 1.116 0.967 1.288 1.500 0.134 Chin-Shan Cohort(20) pooled 1.248 1.000 1.558 1.960 0.050 Fram Offspring (10) men 1.110 0.970 1.270 1.518 0.129 Fram Offspring (10) women 1.200 0.988 1.457 1.840 0.066 AMORIS(21) pooled 1.400 1.356 1.446 20.583 0.000 INTERHEART(22) pooled 1.280 1.246 1.315 17.759 0.000 Women's Health(12) women 1.219 1.128 1.317 5.007 0.000 ISIS(23) pooled 1.487 1.402 1.577 13.210 0.000 Overall 1.249 1.177 1.326 7.311 0.000 0.5 1 2 anti-atherogenic atherogenic Meta Analysis Figure A1. Forest plot of LDL-C relative risk ratios among 13 studies in 11 reports with both apob and non-hdl-c relative risk ratios. The relative risk ratio was significantly greater (p<0.05) than 1.0 within 9 of the 13 studies and highly significant (p<0.001) overall. Heterogeneity was also highly significant (p <0.001). 9

Non-HDL C relative risk ratio (RRR) Study name Subgroup within study Statistics for each study Hazard ratio and 95% CI Hazard Lower Upper ratio limit limit Z-Value p-value Health Professionals (16) men 1.393 1.092 1.778 2.668 0.008 Nurses' Health(11) women 1.572 1.300 1.900 4.670 0.000 MONICA/KORA(17) men 1.485 1.260 1.750 4.718 0.000 MONICA/KORA(17) women 1.794 1.400 2.300 4.617 0.000 Health Professionals (18) men 1.435 1.188 1.734 3.746 0.000 Casale Monferrato(19) pooled 0.910 0.785 1.057-1.235 0.217 Copenhagen Heart(13) men 1.344 1.186 1.522 4.634 0.000 Copenhagen Heart(13) women 1.149 0.995 1.326 1.897 0.058 Chin-Shan Cohort(20) pooled 1.276 1.000 1.629 1.960 0.050 Fram Offspring (10) men 1.218 1.060 1.400 2.781 0.005 Fram Offspring (10) women 1.286 1.060 1.560 2.551 0.011 AMORIS(21) pooled 1.495 1.460 1.530 33.636 0.000 INTERHEART(22) pooled 1.204 1.170 1.240 12.551 0.000 Women's Health(12) women 1.391 1.269 1.523 7.085 0.000 ISIS(23) pooled 1.449 1.375 1.526 13.883 0.000 Overall 1.337 1.239 1.442 7.508 0.000 0.5 1 2 anti-atherogenic atherogenic Meta Analysis Figure A2. Forest plot of non-hdl-c relative risk ratios among 15 independent studies in 12 reports with both apob and non-hdl-c relative risk ratios. The relative risk ratio was significantly greater (p<0.05) than 1.0 within 12 of the 15 studies and highly significant (p<0.001) overall. Heterogeneity was also highly significant (p <0.001). 10

ApoB relative risk ratio (RRR) Study name Subgroup within study Statistics for each study Hazard ratio and 95% CI Hazard Lower Upper ratio limit limit Z-Value p-value Health Professionals (16) men 1.410 1.096 1.814 2.675 0.007 Nurses' Health(11) women 1.800 1.486 2.180 6.016 0.000 MONICA/KORA(17) men 1.490 1.249 1.778 4.422 0.000 MONICA/KORA(17) women 1.730 1.319 2.269 3.963 0.000 Health Professionals (18) men 1.480 1.225 1.788 4.068 0.000 Casale Monferrato(19) pooled 1.170 1.012 1.353 2.121 0.034 Copenhagen Heart(13) men 1.360 1.201 1.541 4.832 0.000 Copenhagen Heart(13) women 1.208 1.019 1.432 2.176 0.030 Chin-Shan Cohort(20) pooled 1.430 1.138 1.797 3.070 0.002 Fram Offspring (10) men 1.370 1.198 1.567 4.592 0.000 Fram Offspring (10) women 1.380 1.145 1.663 3.384 0.001 AMORIS(21) pooled 1.510 1.471 1.551 30.484 0.000 INTERHEART(22) pooled 1.320 1.281 1.361 17.951 0.000 Women's Health(12) women 1.400 1.276 1.536 7.132 0.000 ISIS(23) pooled 1.630 1.538 1.728 16.462 0.000 Overall 1.428 1.347 1.514 11.981 0.000 0.5 1 2 anti-atherogenic atherogenic Meta Analysis Figure A3. Forest plot of apob relative risk ratios among 12 independent published reports with both apob and non-hdl-c relative risk ratios. The relative risk ratio is significantly greater (p<0.05) than 1.0 within each study and highly significant (p<0.001) overall. Heterogeneity was also highly significant (p <0.001). 11

Log apob relative risk ratio (RRR) - non-hdl C RRR Study name Subgroup within study Statistics for each study Point estimate and 95% CI Point Standard Lower Upper estimate error Variance limit limit Z-Value p-value Health Professionals (16) men 0.015 0.041 0.002-0.064 0.094 0.370 0.711 Nurses' Health(11) women 0.118 0.069 0.005-0.017 0.253 1.716 0.086 MONICA/KORA(10) men 0.000 0.028 0.001-0.055 0.055 0.000 1.000 MONICA/KORA(10) women -0.034 0.043 0.002-0.119 0.051-0.783 0.434 Health Professionals(18) men 0.029 0.036 0.001-0.042 0.100 0.805 0.421 Casale Monferrato(19) pooled 0.247 0.076 0.006 0.098 0.396 3.253 0.001 Copenhagen Heart(13) men 0.012 0.020 0.000-0.027 0.051 0.599 0.549 Copenhagen Heart(13) women 0.076 0.029 0.001 0.020 0.132 2.649 0.008 Chin-Shan Cohort(20) pooled 0.116 0.105 0.011-0.090 0.322 1.103 0.270 Fram Offspring (10) men 0.116 0.022 0.000 0.073 0.159 5.228 0.000 Fram Offspring (10) women 0.075 0.031 0.001 0.015 0.135 2.434 0.015 AMORIS(21) pooled 0.007 0.004 0.000-0.001 0.015 1.622 0.105 INTERHEART(22) pooled 0.087 0.005 0.000 0.078 0.096 18.009 0.000 Women's Health(12) women 0.007 0.015 0.000-0.022 0.036 0.471 0.638 ISIS(23) pooled 0.118 0.009 0.000 0.100 0.136 12.593 0.000 Overall 0.056 0.016 0.000 0.024 0.087 3.425 0.001-0.25-0.13 0.00 0.13 0.25 non-hdl C <- favors->apob Meta Analysis Figure A4. Log apob RRR log non-hdl-c RRR forest plot of 12 independent published studies with both apob and non-hdl-c vascular risk associations. This analysis counts the three reports stratified by sex as six separate studies and excludes a Women s Health Study report(12) based on a subset of the later Women s Health Study report (19). The overall mean difference in log relative risk ratios (95% confidence interval) of 0.056 (0.024, 0.087) equates to a 5.7% (2.4%, 9.1%), p<0.001, advantage favoring apob over non-hdl-c as a marker of atherogenic risk. 12 12

Log non-hdl C relative risk ratio (RRR) - log LDL C RRR Study name Subgroup within study Statistics for each study Point estimate and 95% CI Point Standard Lower Upper estimate error Variance limit limit Z-Value p-value Health Professionals(16) men 0.115 0.037 0.001 0.042 0.188 3.104 0.002 Nurses' Health(11) women 0.134 0.044 0.002 0.047 0.221 3.009 0.003 Health Professionals(18) men 0.101 0.054 0.003-0.004 0.207 1.888 0.059 Casale Monferrato(19) pooled 0.010 0.026 0.001-0.041 0.061 0.386 0.700 Copenhagen Heart(13) men 0.051 0.024 0.001 0.005 0.097 2.158 0.031 Copenhagen Heart(13) women 0.004 0.025 0.001-0.046 0.053 0.142 0.887 Chin-Shan Cohort(20) pooled 0.022 0.026 0.001-0.029 0.074 0.849 0.396 Fram Offspring (10) men 0.094 0.024 0.001 0.047 0.142 3.889 0.000 Fram Offspring (10) women 0.065 0.034 0.001-0.003 0.132 1.885 0.059 AMORIS(21) pooled 0.055 0.007 0.000 0.042 0.068 8.378 0.000 INTERHEART(22) pooled -0.056 0.005 0.000-0.066-0.046-11.120 0.000 Women's Health(12) women 0.132 0.016 0.000 0.100 0.164 8.049 0.000 ISIS(23) pooled -0.026 0.010 0.000-0.046-0.005-2.466 0.014 Overall 0.050 0.021 0.000 0.009 0.091 2.380 0.017-0.25-0.13 0.00 0.13 0.25 LDL C <- favors -> non-hdl C Meta Analysis Figure A5. Log non-hdl-c RRR log LDL-C RRR forest plot of the 11 independent published reports with both non-hdl-c and LDL-C relative risk ratios among the 13 published reports with both apob and non-hdl-c relative risk ratios. This analysis counts the two reports stratified by sex as four separate studies and excludes a Women s Health Study report(12) based on a subset of the later Women s Health Study report (19). The overall mean difference in log relative risk ratios (95% confidence interval) of 0.050 (0.009, 0.091) equates to a 5.1% (0.9%, 9.5%), p=0.017, advantage favoring non-hdl-c over LDL-C as a marker of atherogenic risk. 13

Log apob relative risk ratio (RRR) - log LDL C RRR Study name Subgroup within study Statistics for each study Point estimate and 95% CI Point Standard Lower Upper estimate error Variance limit limit Z-Value p-value Health Professionals (16) men 0.125 0.051 0.003 0.025 0.225 2.440 0.015 Nurses' Health(11) women 0.251 0.066 0.004 0.122 0.380 3.824 0.000 Health Professionals(18) men 0.130 0.055 0.003 0.022 0.239 2.353 0.019 Casale Monferrato(19) pooled 0.257 0.035 0.001 0.188 0.326 7.332 0.000 Copenhagen Heart(13) men 0.063 0.038 0.001-0.011 0.137 1.663 0.096 Copenhagen Heart(13) women 0.079 0.041 0.002-0.002 0.161 1.918 0.055 Chin-Shan Cohort(20) pooled 0.138 0.100 0.010-0.058 0.334 1.384 0.166 Fram Offspring (10) men 0.210 0.043 0.002 0.125 0.296 4.847 0.000 Fram Offspring (10) women 0.140 0.055 0.003 0.032 0.248 2.538 0.011 AMORIS(21) pooled 0.061 0.008 0.000 0.047 0.076 8.167 0.000 INTERHEART(22) pooled 0.031 0.007 0.000 0.017 0.045 4.363 0.000 Women's Health(12) women 0.139 0.022 0.000 0.097 0.182 6.402 0.000 ISIS(23) pooled 0.093 0.016 0.000 0.061 0.125 5.671 0.000 Overall 0.120 0.018 0.000 0.085 0.154 6.823 0.000-0.25-0.13 0.00 0.13 0.25 LDL C <- favors -> apob C Meta Analysis Figure A6. Log apob RRR log LDL-C RRR forest plot of the 11 independent published reports with both apob and LDL-C relative risk ratios among the 13 reports with both apob and non-hdl-c relative risk ratios. This analysis counts the two reports stratified by sex as four separate studies and excludes a Women s Health Study report(12) based on a subset of the later Women s health Study report (19). The overall mean difference in log relative risk ratios (95% confidence interval) of = 0.120 (0.085, 0.154) = equates to a 12.7% (8.9%, 16.6%), p<0.001, advantage apob over LDL-C as a marker of atherogenic risk. 14

Figure A7. Subgroup analyses by study attributes. The p-values for each section indicate result from testing equality of the difference between the apob and non-hdl-c log relative risk ratios across the section s subgroups. Assay well documented reflects the authors judgment regarding reported assurances re: assay/specimen quality. Mean of subgroup comparison versus 0%: *p<0.05, ** p<0.01, ***p<0.001 % 15

A 500 B C 200 A. ApoB versus NonHDL-C B. NonHDL-C versus LDL-C C. ApoB versus LDL-C Precision (1/standard error) 0.055 (0.024, 0.087) 0.047 (0.016, 0.078) 0.050 (0.009, 0.091) 0.044 (0.005, 0.084) 0.120 (0.085, 0.154) 0.058 (0.024, 0.093) Figure A8. Selection bias assessments: funnel plots of precision by difference in log relative risk ratios. The empty circles represent the observed studies while the empty diamonds represents the overall random effects means and 95% confidence intervals of the observed studies. If smaller studies were not published owning to non-significant findings, one would expect the distribution of empty circles to be skewed to the right. The black filled circles represent studies imputed to balance the distribution around the overall estimate and the black diamonds represent overall average adjusted by including the imputed studies. After adjustment each overall difference remains statistically significant. 16