Abstract/Summary

Public dietary guidelines worldwide recommend that dietary saturated fatty acids (SFAs) do not exceed 10% total energy (%TE), and their replacement with unsaturated fatty acids (UFAs) has been shown to help prevent cardiometabolic disease (CMD). However, research on the impact of dietary SFAs on CMD risk mostly relies on indirect evidence either from intervention studies measuring biomarkers of disease risk, such as fasting lipid profiles, or from observational prospective cohort studies with disease outcomes. In addition, emerging research suggests the impact of dietary SFAs on cardiometabolic health might be modulated by novel CMD risk markers identified by omics approaches, differential effects of individual SFAs, and/or specific food matrix effects. In particular, dairy products contribute to 21% of dietary SFAs intakes in UK adults, but their consumption does not seem to be associated with increased CMD risk according to epidemiological studies. In this context, this PhD thesis aimed to (i) investigate the impact of overall and individual dietary SFAs on medium�term CMD risk markers and long-term CMD risk, and (ii) assess the utility of dairy foods, and more particularly dairy fat, for CMD prevention at a population level. I first assessed the plasma lipidome-mediated impact of isoenergetically replacing dietary SFAs with monounsaturated fatty acids (MUFAs) or a mixture of MUFAs and polyunsaturated fatty acids (PUFAs) on CMD risk markers and long-term CMD risk (Chapter 2). To achieve this, joint lipidomics analyses in a subset of n=113 participants from the DIVAS randomised controlled trial (RCT) and a sub-cohort from the EPIC-Potsdam prospective cohort study (specific case-cohorts: n=1,707 and n=775 cases for type 2 diabetes, n=1,886 and n=551 cases for cardiovascular disease [CVD]) were completed. This secondary analysis showed that UFA-rich diets implemented over 16 weeks to reduce dietary SFAs in the DIVAS RCT significantly reduced the plasma concentrations of SFA-containing glycerolipids (i.e. mono-, di-, and triacylglycerols) and sphingolipids which were associated with long-term CVD risk in the EPIC-Potsdam cohort study. In addition, I identified that increased serum concentrations of low�density lipoprotein cholesterol (LDL-C), an established CMD risk marker, were associated with higher plasma levels of glycerolipids containing lauric (12:0) and stearic acids (18:0). The impact of individual SFAs was further assessed at the dietary level by conducting a systematic literature review and meta-analysis of 44 RCTs which substituted individual dietary SFAs with another fatty acid (FA) or a mixture of UFAs (Chapter 3). In quantitative meta-analyses, I observed reductions in LDL-C concentrations after the replacement of palmitic acid (C16:0) with UFAs (-0.36 mmol/L, 95%CI -0.50 to -0.21, I2=96.0%, n=18 RCTs) or oleic acid (C18:1) (-0.16 mmol/L, 95% CI -0.28 to -0.03, I2=89.6%, n=9 RCTs), with a similar impact on total cholesterol and apolipoprotein B concentrations. Furthermore, I identified important research gaps regarding the impact of individual dietary SFAs on novel CMD risk markers (e.g. markers of inflammation, endothelial activation, and glycaemic control) and the specific effect of short-chain SFAs, lauric acid (12:0), and myristic acid (14:0). To assess the practical applications of reducing dietary SFAs in free-living UK adults, we developed a food-based dietary fat exchange model (the RISSCI-1 study), which aimed to replace dietary SFAs with UFAs by replacing high-fat dairy and high-SFA snacks with commercially available lower-fat dairy foods along with high-UFA cooking oil and snacks into the habitual diets of n=109 UK adult men for 4 weeks (Chapter 4). Participants successfully exchanged 10.4%TE of dietary SFAs with 9.7%TE UFAs, with minimal impact on other nutrients. In addition, participants incorporated the intervention food items without changing their overall dietary habits. Importantly, the analysis of plasma phospholipid fatty acids (PL FAs) in the RISSCI-1 study, along with those performed in the SATgen, DIVAS, and RESET dietary fat intervention studies (Chapter 5), revealed contrasted results on the validity of individual plasma PL FAs as proxies for dairy consumption. In particular, circulating odd-chain SFAs and ruminant trans FAs have been commonly use as biomarkers of intakes in epidemiological studies so far, but only modestly correlate with overall dairy intakes in RCTs and do not seem to accurately capture the intakes of low-fat dairy foods (Chapter 1). Finally, a prospective analysis of the NutriNet-Santé cohort study did not reveal statistically significant associations between overall and specific dairy consumption and overall CVD (n=1,952 cases) or coronary heart disease risk (n=1,219 cases) among n=104,805 French adults (Chapter 6). However, we observed a 19% reduction (HR=0.81, 95%CI 0.66 to 0.98, p-trend=0.01) in cerebrovascular disease risk (n=878 cases) associated with higher intakes (i.e. at least 160 g/d) of fermented dairy foods (i.e. yogurt, cheese, and fermented milk) compared to low intakes (i.e. below 57 g/d). Despite being observational, these results generated new hypotheses on the potential beneficial effects of specific dairy food matrices on CMD risk, which may stem from bioactive peptides, calcium, and the fermentation process. Overall, results from this PhD thesis concur with current dietary guidelines on the reduction of dietary SFAs in favour of UFAs and suggest that their deleterious impact on CMD risk may be modulated by their individual structure and/or their effect on the plasma lipidome. Moreover, these findings provide further evidence supporting the presence of beneficial compounds within the dairy food matrix, which may counterbalance the potential deleterious effects of their SFA content. Nonetheless, further interventional and observational studies are warranted to validate these conclusions, and future research is needed to elucidate the physiological mechanisms underlying the complex interactions between dietary SFAs, the dairy food matrix, and the physiological response to dairy consumption.