We determined the effect of consumption of an HFD and exercise training performed in the morning or in the evening on circulating lipoprotein subfractions in men with overweight/obesity. We report that 5 days of HFD induced substantial changes in lipoprotein subfraction profiles, with alterations in VLDL, IDL, LDL, and HDL subfractions. Based on previous evidence of associations between lipoprotein subfractions and risk of cardiometabolic diseases6,7,16,17,18,19,20,21, we interpret the effect of the HFD on lipoprotein subfractions as mostly beneficial for cardiometabolic health. Daily exercise training for 5 days also induced distinct and favourable changes in lipoprotein subfraction profiles, with no clear difference between morning and evening exercise.
Consumption of an HFD for 5 days reduced the total fasting concentrations of cholesterol, free cholesterol, and phospholipids in VLDL, as well as cholesterol and triglycerides in several VLDL subfractions. Higher levels of large VLDL particles are associated with insulin resistance16 and incident type 2 diabetes17, independent of established risk factors such as circulating glucose and insulin concentrations. Furthermore, cholesterol in VLDL explained 40% of the excess risk of myocardial infarction associated with obesity in 29,010 individuals from the Copenhagen General Population Study18. In our study, consumption of the HFD primarily impacted the larger VLDL particles (VLDL-1–3), with no significant changes in VLDL-4 and VLDL-5. Fasting concentrations of VLDL-1 phospholipids were decreased after the HFD, with a tendency (q = 0.060) of decreased free cholesterol in VLDL-1. Streese and colleagues reported that both VLDL-1 phospholipids and free cholesterol concentrations were inversely associated with retinal arterio-to-venular diameter ratio, an independent measure of cardiovascular outcomes19. There were no effects of the HFD on main VLDL composition in postprandial samples after the HFD, but some of the subfraction components changed (Fig. 3). The timing of blood collection after a meal has implications on VLDL concentrations22, but in our study the timing of these collections was standardised on measurement days. We found that exercise training undertaken in the morning, but not in the evening, further decreased the concentrations of VLDL-1 free cholesterol, phospholipids, and triglycerides. There was also a tendency (q = 0.055) of reduced total triglycerides in VLDL after morning exercise training. The beneficial effect of exercise training on large VLDL particles is in agreement with a recent study reporting an inverse correlation between cardiorespiratory fitness (maximum oxygen uptake) and several VLDL subfractions (e.g., VLDL-1 free cholesterol, VLDL-1 phospholipids, and VLDL-1 triglycerides)23. It has been suggested that high intensity exercise is necessary to decrease the hepatic release of VLDL, and lower VLDL concentrations were only associated with a moderate-to-high intensity, and not low-intensity, aerobic physical activity in 509 participants with increased risk of impaired glucose regulation in the Walking Away from Diabetes study24. The exercise training protocol in the current study included three high-intensity interval training sessions, which may explain the beneficial effect of exercise on VLDL subfractions.
The increase in total serum cholesterol concentrations in LDL after 5 days of HFD was due to increased concentrations of cholesterol in the larger LDL subfractions, with no changes in small, dense LDL particles. We found no increase in the fasting concentrations of Apo-B, an indirect measure of LDL particle number, in total serum or in LDL. There was increased concentrations of Apo-B in larger LDL particles (LDL-2 and LDL-3), again revealing a shift towards larger, more buoyant LDL particles after the HFD. These findings are in line with a systematic review and meta-analysis showing an overall trend for an increase in the larger subclasses of LDL and a decrease in the smaller, more dense LDL subclasses, after carbohydrate-restricted dietary interventions11. The circulation time of smaller LDL particles is longer than that of large LDL particles, with small dense LDL particles being more susceptible to atherogenic modifications, including glycation and oxidation25. Indeed, several studies have shown a strong association between small dense LDL particles and incidence of cardiovascular disease6,20,21. For example, larger LDL particles showed no association with future coronary heart disease events in the Atherosclerosis Risk in Communities (ARIC) study, whereas the concentration of small-density LDL cholesterol predicted later incidence of such disease, independent of traditional cardiovascular risk factors in this large cohort study with 11,419 participants21.
We found that exercise training mostly affected large LDL particles, with reductions in fasting Apo-B, cholesterol, free cholesterol, and phospholipids concentrations in LDL-1 after both morning and evening exercise. There were additional reductions in some LDL-3 and LDL-4 subfractions after evening exercise training. These findings, in part, contrast with the results of a meta-analysis of 10 exercise interventions lasting 20–26 weeks which showed that large LDL particles increased and small dense LDL particles decreased after endurance exercise training26. The reason for these discrepant findings may be that the exercise intervention period in our study was short term (5 days).
HDL particles are heterogeneous in their size and composition and standard clinical measurement of HDL cholesterol concentrations is unable to capture this diversity. The findings of recent studies indicate that the inverse associations of cholesterol in HDL particles and incident type 2 diabetes and ASCVD are limited to large and medium subclasses7,27,28,29,30. In fact, some studies show that concentrations of smaller HDL particles are associated with higher risk of developing type 2 diabetes28,30. Diets high in fat (41–62% of total energy intake, TEI) generally increase total HDL cholesterol31, but this effect may be mediated by a concomitant reduction in body mass32. In our study, there was no change in total body mass and no change in total HDL cholesterol after 5 days of HFD. However, the HFD induced significant reductions in several HDL-related variables. There was reduced triglyceride enrichment in HDL-3 and HDL-4, indicating a beneficial effect of HFD since triglyceride concentrations in these small HDL particles are associated with risk of myocardial infarction7. Moreover, triglycerides concentration in HDL, most prominently HDL-3, is associated with reduced microvascular health (retinal arterio-to-venular diameter ratio)19 and low cardiorespiratory fitness23, both of which are important predictors of ASCVD events and mortality33,34,35.
Exercise training also primarily affected the smallest HDL particles, producing reduced fasting concentrations of Apo-A1, Apo-A2, and free cholesterol in both HDL-3 and HDL-4, and additionally lower HDL-4 phospholipids concentrations, after both morning and evening exercise. These findings are in agreement with a previous study that reported reduced concentrations of small HDL particles after 4 days of daily exercise (20 min of moderate intensity endurance exercise) in sedentary but otherwise healthy men, despite no change in total HDL concentrations36.
Strengths of the current study include its randomised design, rigorous dietary control with all meals provided to the participants, along with prescribed times for eating, and supervised exercise sessions in the laboratory (i.e., little uncontrolled environmental stimuli). The comprehensive analyses of lipoprotein particle profile quantified using NMR spectroscopy, with examination of different lipoprotein attributes not measured in a standard lipid profile, is a major strength of our study. However, there is no standardised method to analyse lipoprotein subfractions, with different methods utilising different techniques for separation of subfractions, making it challenging to directly compare our results with those of others. The timing of blood collection post-exercise can affect circulating concentrations of lipoproteins. For example, decreased total plasma triglyceride concentrations and increased concentrations of cholesterol in HDL were reported 24 h after exercise, and lasted through 48 h, after a single session of moderate-intensity treadmill walking in physically inactive men37. Due to the design of our study, which aimed to compare the effect of morning versus evening exercise, there was a difference in the timing of biological sampling since the last exercise session (12 vs 24 h for fasting samples and 24 vs 36 h for postprandial samples). This is a limitation in our study and such discrepancy is inherent in any investigation of the effects of the time of day of exercise. We only included men in our study and the results cannot be generalised to women. Furthermore, the small sample size per group may limit the interpretation of the findings due to large variability between individuals for some of the variables.
Our results are from a short-term experimental intervention albeit with strict control of the participants’ dietery intake, and accordingly, our findings should not be taken as clinical recommendations. We acknowledge that clinical guidelines for primary prevention of ASCVD from the major cardiology associations (ESC and AHA/ACC), recommend to reduce the intake of saturated fats38,39. However, several meta-analyses and recent studies indicate that there is no clear association between intake of total fat, or saturated fat, and risk of ASCVD8,9,10,40. Indeed, there is a current spirited debate in the scientific community about the scientific basis for the concept that fat in general, and saturated fat in particular, causes ASCVD41,42.
Short-term consumption of a HFD in men with overweight/obesity induced marked alterations in lipoprotein subfraction profiles, including reductions in several large VLDL subfractions and reduced triglycerides concentrations in small HDL particles. Daily exercise while consuming the HFD, undertaken either in the morning or in the evening, resulted in a distinctive lipoprotein subfraction signature, compared with no exercise. The effect of exercise was particularly evident for the largest LDL particles, with lower concentrations of Apo-B, cholesterol, free cholesterol, and phospholipids in LDL-1, as well as for several subfractions in the smallest HDL particles. Taken collectively, we interpret the overall effects of both the HFD and exercise intervention on lipoprotein subfractions as beneficial for cardiovascular disease prevention.