Pattern of lipid biomarkers and risk of cardiovascular disease Robert Clarke Clinical Trial Service Unit University of Oxford 9 January 2017
Biomarkers for dietary fats Blood lipids (LDL, HDL, triglycerides, ApoB, ApoA 1 ) Specific fatty acids Polyunsaturated fatty acids (PUFA) Monounsaturated fatty acids (MUFA) Saturated fatty acids (SAFA) Omega-3 fatty acids (EPA & DHA)
Ancel Keys (1904-2004) Keys advocated that replacing saturated by unsaturated fat would reduce blood cholesterol and the incidence of CHD. The resulting changes in the composition of food fats led to a doubling of the proportion of the unsaturated fatty acid, linoleic acid, in USA between 1960 and 1975. Rates of CHD in the US began to fall from 1968 onwards, and about half of the decline was ascribed to changes in lifestyle including diet.
Seven Countries Study: SFA, cholesterol and CHD mortality Saturated fat and risk of CHD mortality Total cholesterol and risk of CHD mortality
Confusion about reliability of evidence on diet and disease Prospective studies of both dietary fatty acids and fatty acid biomarkers show no association of SAFA, PUFA or MUFA with risk of coronary heart disease, but positive association with TFA and weak inverse association with n-3 FA Chowdhury, Ann Int Med 2014 The limitations of such studies have not been widely appreciated prompting confusion about the optimum amount and type of dietary fat
Confusion about amount and type of dietary fat Consensus Harms of trans fats Harms of processed meat Lack of effect of total fat Insufficient evidence Coconut oil Palm oil Whole fat milk Some uncertainty N-3 PUFA N-6 PUFA Plant derived MUFA Substantial uncertainty SAFA Dietary cholesterol Low fat milk Unprocessed red meat Mozaffarian, Circulation 2016
Criteria for an informative biomarker Biomarker is on the causal pathway between diet and disease Significant association of biomarker with disease in the target population Biomarker changes consistently with the end-point in response to intervention Change in the biomarker explains a substantial proportion of the change in response to the intervention
Hazard ratio (floating absolute risks & 95% CI) CHD mortality versus total CHOLESTEROL: Prospective Studies Collaboration 33,744 deaths at ages 40-89 256 Age at risk 80-89 1 mmol/l lower total cholesterol 15% lower risk 128 70-79 18% lower risk 64 60-69 28% lower risk 32 50-59 42% lower risk 16 40-49 56% lower risk 8 4 2 1 0 5 4 0 5 0 6 0 7 0 8 0 Usual total cholesterol (mmol/l) Prospective Studies Collaboration, Lancet 2007
CHD mortality versus non-hdl and HDL-cholesterol: Prospective Studies Collaboration Positive association between non-hdl and risk of CHD Inverse association between HDL and risk of CHD Prospective Studies Collaboration; Lancet 2007
Metabolic ward studies Metabolic ward studies are dietary trials conducted over several weeks in which all food consumed is provided Participants have a fixed caloric intake and the amount and type of fat consumed is recorded & lipids measured after diet Meta-analysis included 444 dietary experiments with total cholesterol and 227 with LDL-C Unpublished
Effects of SFA on total cholesterol & LDL-C Total cholesterol LDL-C
Effects of replacing carbohydrate by fats on LDL-C versus HDL-C Saturated fat (% calories) Effect on LDL-C (mmol/l) n=227 0.035 (0.026, 0.044) Effect on HDL-C (mmol/l) n=245 0.010 (0.007, 0.013) Polyunsaturated fat (% calories) -0.026 (-0.036, -0.015) 0.002 (-0.002, 0.005) Monounsaturated fat (% calories) -0.011 (-0.020, -0.002) 0.005 (0.002, 0.008) Dietary cholesterol (100mg/day) 0.057 (0.030, 0.084) -0.04 0 0.04 0.08 0.12 Decrease Increase 9e-04 (-0.009, 0.010) -0.04 0 0.04 0.08 0.12 Decrease Increase
Effects of replacing carbohydrate by fats on LDL-C versus Total/HDL-C ratio Saturated fat (% calories) Effect on LDL-C (mmol/l) n=227 0.035 (0.026, 0.044) Effect on Total/HDL-C ratio n=245 0.007 (-0.002, 0.015) Polyunsaturated fat (% calories) -0.026 (-0.036, -0.015) -0.024 (-0.034, -0.013) Monounsaturated fat (% calories) -0.011 (-0.020, -0.002) -0.020 (-0.029, -0.011) Dietary cholesterol (100mg/day) 0.057 (0.030, 0.084) -0.04 0 0.04 0.08 0.12 Decrease Increase 0.053 (0.026, 0.079) -0.04 0 0.04 0.08 0.12 Decrease Increase
Effect of replacement of 5% calories as SFA by PFA, MFA or carbohydrate Replacement of 5% calories as saturated fat by polyunsaturated fat Tot Chol (mmol/l) LDL-C (mmol/l) HDL-C (mmol/l) Total/HDL-C ratio Tot Chol (mmol/l) Replacement of 5% calories as saturated fat by monounsaturated fat LDL-C (mmol/l) HDL-C (mmol/l) Total/HDL-C ratio Replacement of 5% calories as saturated fat by complex carbohydrate Tot Chol (mmol/l) LDL-C (mmol/l) HDL-C (mmol/l) Total/HDL-C ratio -0.4-0.3-0.2-0.1 0.0 0.1 Decrease Increase
Lessons from metabolic ward studies Replacement of saturated fat by unsaturates rather than by carbohydrate produces much greater reductions in LDL-C and in total/hdl-c ratio Feasible exchanges of saturated fats (5% calories) and trans fats (2% calories) and dietary cholesterol (100mg) by polyunsaturated fat should lower LDL- C by 0.5 mmol/l and total/hdl-c by 0.33 Promote exchange of foods high in saturated fat, by those with high levels of polyunsaturates and reduce dietary cholesterol and eliminate trans fats
Large trials of dietary intake of PUFA (15% vs 5%) for prevention of major vascular events Mozaffarian, PLoS One 2010
P r o p o r t i o n a l r e d u c t i o n i n r i s k Reduction in risk of major vascular events is proportional to differences in LDL: Cholesterol Treatment Trialists meta-analysis 5 0 % 4 0 % 3 0 % 2 0 % 1 0 % 0 0. 5 1. 0 1. 5 2. 0 R e d u c t i o n i n L D L c h o l e s t e r o l ( m m o l / L ) However, trials of HDL-raising drugs have not yet shown benefits for CHD
Large trials of omega-3 fatty acids for prevention of Major Vascular Events (10 trials, 78,000 participants) Submitted
China Kadoorie Biobank study Conventional blood-based lipid measurements (e.g. LDL-C, HDL-C, TG) crudely quantify lipids A nested case control study of MI (912 cases) measured NMRmetabolites on stored plasma 1 H-NMR metabolomics (Brainshake assay) quantified 225 metabolites, 85% lipid related Logistic regression was used to test the associations of metabolites with risk of CHD Analyses were adjusted for age, sex, fasting status, educational attainment, smoking, geographical region
Lipidomics and CHD in the China Kadoorie Biobank study Particle Number Cholesterol Concentration Triglyceride Concentration Unpublished
Conclusions Replacement of saturated fat by unsaturates rather than by carbohydrate produces much greater reductions in LDL-C and in total/hdl-c ratio Large trials confirm benefits for major vascular events of replacement of SAFA by PUFA and of lowering LDL by any means, but no benefit of n-3 FA for prevention of major vascular events LDL remains the optimum biomarker to assess the effects of dietary fats and particle number or size add little beyond LDL