Thomas Olsen1, Kathrine J Vinknes2, Gard F T Svingen3, Eva R Pedersen3, Grethe S Tell4, Rune Blomhoff5, Christian A Drevon2, Per M Ueland6, Øivind Midttun7, Helga Refsum2, Ottar K Nygård8. 1. Department of Nutrition, Institute of Basic Medical Sciences, Faculty of Medicine, University of Oslo, 0372 Oslo, Norway. Electronic address: thool@uio.no. 2. Department of Nutrition, Institute of Basic Medical Sciences, Faculty of Medicine, University of Oslo, 0372 Oslo, Norway. 3. Department of Heart Disease, Haukeland University Hospital, 5021 Bergen, Norway; KG Jebsen Centre for Diabetes Research, University of Bergen, 5009 Bergen, Norway. 4. Department of Global Public Health and Primary Care, University of Bergen, 5018 Bergen, Norway. 5. Department of Nutrition, Institute of Basic Medical Sciences, Faculty of Medicine, University of Oslo, 0372 Oslo, Norway; Division of Cancer Medicine, Oslo University Hospital, 0450 Oslo, Norway. 6. Laboratory of Clinical Biochemistry, Haukeland University Hospital, 5021 Bergen, Norway; Bevital AS, 5021 Bergen, Norway. 7. Bevital AS, 5021 Bergen, Norway. 8. Department of Heart Disease, Haukeland University Hospital, 5021 Bergen, Norway; KG Jebsen Centre for Diabetes Research, University of Bergen, 5009 Bergen, Norway; Department of Clinical Science, Faculty of Medicine, University of Bergen, 5021 Bergen, Norway.
Abstract
BACKGROUND AND AIMS: Apolipoproteins B (apoB) and A1 (apoA1) are major protein constituents of low-density and high-density lipoproteins, respectively, and serum concentrations of these apolipoproteins are associated with risk of atherosclerosis. Vitamin A (VA) has been implicated in lipoprotein metabolism. We evaluated the associations of serum apoB, apoA1 and their ratio (apoBAR) with risk of incident acute myocardial infarction (AMI) and the possible modification by serum VA. METHODS: Risk associations were assessed by Cox regression, and presented as hazard ratios (HRs) per standard deviation (SD) increment in log-transformed values of the lipid parameters, among 4117 patients with suspected stable angina pectoris, located in Western Norway. Interactions with VA were evaluated by including interaction terms in the models. The multivariate model included age, sex, smoking, hypertension, number of stenotic coronary arteries, left ventricular ejection fraction, C-reactive protein, estimated glomerular filtration rate and statin treatment at discharge. RESULTS: Median (25th, 75th percentile) age of the 4117 patients (72% male) was 62 (55, 70) years. ApoB and apoA1 were higher among patients in the upper versus lower tertiles of VA. During a median of 4.6 (3.6, 5.7) years of follow-up, 8.2% of patients experienced an AMI. Overall, we observed no significant associations between lipid parameters and AMI after multivariate adjustment. However, apoB and apoBAR were associated with AMI among patients in the upper tertile of VA (HR per SD 1.35, (95% CI: 1.11-1.65), and 1.42 (1.16-1.74), respectively, p for interactions ≤0.003). CONCLUSIONS: The associations of apoB and apoBAR with incident AMI were confined to patients with elevated VA.
BACKGROUND AND AIMS: Apolipoproteins B (apoB) and A1 (apoA1) are major protein constituents of low-density and high-density lipoproteins, respectively, and serum concentrations of these apolipoproteins are associated with risk of atherosclerosis. Vitamin A (VA) has been implicated in lipoprotein metabolism. We evaluated the associations of serum apoB, apoA1 and their ratio (apoBAR) with risk of incident acute myocardial infarction (AMI) and the possible modification by serum VA. METHODS: Risk associations were assessed by Cox regression, and presented as hazard ratios (HRs) per standard deviation (SD) increment in log-transformed values of the lipid parameters, among 4117 patients with suspected stable angina pectoris, located in Western Norway. Interactions with VA were evaluated by including interaction terms in the models. The multivariate model included age, sex, smoking, hypertension, number of stenotic coronary arteries, left ventricular ejection fraction, C-reactive protein, estimated glomerular filtration rate and statin treatment at discharge. RESULTS: Median (25th, 75th percentile) age of the 4117 patients (72% male) was 62 (55, 70) years. ApoB and apoA1 were higher among patients in the upper versus lower tertiles of VA. During a median of 4.6 (3.6, 5.7) years of follow-up, 8.2% of patients experienced an AMI. Overall, we observed no significant associations between lipid parameters and AMI after multivariate adjustment. However, apoB and apoBAR were associated with AMI among patients in the upper tertile of VA (HR per SD 1.35, (95% CI: 1.11-1.65), and 1.42 (1.16-1.74), respectively, p for interactions ≤0.003). CONCLUSIONS: The associations of apoB and apoBAR with incident AMI were confined to patients with elevated VA.
Authors: Thomas Olsen; Kathrine J Vinknes; Rune Blomhoff; Vegard Lysne; Øivind Midttun; Indu Dhar; Per M Ueland; Gard F T Svingen; Eva K R Pedersen; Christian A Drevon; Helga Refsum; Ottar K Nygård Journal: Eur J Nutr Date: 2019-09-09 Impact factor: 5.614