Abstract
Individuals with discordantly high apoB to LDL-cholesterol levels carry a higher risk of atherosclerotic CVD compared with those with average or discordantly low apoB to LDL-cholesterol. We aimed to determine associations between apoB and LDL-cholesterol discordance in relation to nutrient patterns (NP) using National Health and Nutrition Examination Survey data. Participants were grouped by established LDL-cholesterol and apoB cut-offs (Group 1: low apoB/low LDL-cholesterol, Group 2: low apoB/high LDL-cholesterol, Group 3: high apoB/low LDL-cholesterol, Group 4: high apoB/high LDL-cholesterol). Principle component analysis was used to define NP. Machine learning (ML) and structural equation models were applied to assess associations of nutrient intake with apoB/LDL-cholesterol discordance using the combined effects of apoB and LDL-cholesterol. Three NP explained 63·2 % of variance in nutrient consumption. These consisted of NP1 rich in SFA, carbohydrate and vitamins, NP2 high in fibre, minerals, vitamins and PUFA and NP3 rich in dietary cholesterol, protein and Na. The discordantly high apoB to LDL-cholesterol group had the highest consumption of the NP1 and the lowest consumption of the NP2. ML showed nutrients that had the greatest unfavourable dietary contribution to individuals with discordantly high apoB to LDL-cholesterol were total fat, SFA and thiamine and the greatest favourable contributions were MUFA, folate, fibre and Se. Individuals with discordantly high apoB in relation to LDL-cholesterol had greater adherence to NP1, whereas those with lower levels of apoB, irrespective of LDL-cholesterol, were more likely to consume NP3.
Highlights
A high concentration of fasted serum LDL-cholesterol is an independent causal risk factor for atherosclerotic CVD (ASCVD)(1)
The application of machine learning (ML) for the evaluation of nutrient patterns (NP) offers advantages compared with traditional techniques, including the identification of novel trends and patterns derived from models built and operated without human intervention or assistance, unlike traditional statistical approaches where human intervention is essential at every stage of the model-building process[18]
The aim of the present study was to apply two methods of determining NP to data derived from National Health and Nutrition Examination Survey (NHANES) to reveal, for the first time, the association between our outcome and nutrient intake between four groups stratified by apoB and LDL-cholesterol
Summary
A high concentration of fasted serum LDL-cholesterol is an independent causal risk factor for atherosclerotic CVD (ASCVD)(1). Others may have higher LDL-cholesterol and lower apoB, a profile indicative of a smaller number of LDL particles that are larger in size, which result in a reduced ASCVD risk[3] In this regard, it has been shown that the size of atherogenic lipoprotein particles is more discriminatory, as smaller particles more readily penetrate the endothelial wall and have greater susceptibility to atherogenic modification than their larger counterparts; a process that initiates and drives ASCVD[5]. The risk of ASCVD is more strongly associated with the number of apoB particles rather than their cholesterol content[10,11,12,13,14] This risk is increased by the presence of additional molecules bound to the apoB-containing lipoprotein complex, for example, oxidised phospholipids, which have been shown to have major proatherogenic and proinflammatory roles[15]. Structural equation models (SEM) offer an improved ability to determine complex networks (magnitude of associations), as well as benefitting from multicollinearity and other features not possible with traditional statistical models[19]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
