Replacement of saturated fatty acids from meat by dairy sources in relation to incident cardiovascular disease: The European Prospective Investigation into Cancer and Nutrition (EPIC)-Norfolk study

Background Prospective observational data revealed lower cardiovascular disease (CVD) incidence with modelled replacement of saturated fatty acids (SFA) from total meat by total dairy, but it is unknown what the associations are of replacing SFA from types of meat by types of dairy with CVD incidence. Objective To investigate the associations of replacing SFA from total, red, processed and poultry meat by SFA from total dairy, milk, cheese, and yogurt with the incidence of CVD. Methods We analyzed longitudinal data from 21841 participants of the European Prospective Investigation into Cancer and Nutrition (EPIC)-Norfolk study (56.4% female; age: 40-79 years). Dietary data were collected by food frequency questionnaires at baseline (1993-1997). Incident fatal or non-fatal CVD (n=5902), CHD (n=4215), stroke (total: n=2544; ischaemic: n=1113; haemorrhagic: n=449) were identified up to 2018. Hazard ratios (HR) and 95% confidence intervals (CI) were estimated using Cox regression for the risk associated with the replacement of 2.5% of energy from SFA from meat by dairy, adjusted for sociodemographic, lifestyle, energy, dietary and cardiometabolic factors. Results Replacing SFA from total meat by total dairy was associated with a lower CVD incidence (HR, 0.89; 95% CI, 0.82,0.96) and CHD (0.88; 0.80,0.96). Replacing SFA from processed meat by cheese was associated with lower CVD (0.77; 0.68,0.88); CHD (0.77; 0.66,0.90) and stroke (0.81; 0.67,0.99). Similarly, replacing SFA from red meat by cheese was associated with lower CVD (0.86; 0.76,0.97). Higher incidence of stroke was found with replacement of SFA from poultry by milk (2.06; 1.09,3.89), yogurt (2.55; 1.27,5.13) or cheese (1.96; 1.04,3.70), but the CIs were relatively large due to low, narrow range of poultry SFA intakes. Conclusions Findings indicate that different SFA-rich foods at baseline have differential associations with CVD risk. If confirmed by further studies, these findings could be used to inform specific food-based dietary guidance.