Endocrinology and Metabolism

Effects of moderate exercise on VLDL1 and Intralipid kinetics in overweight/obese middle-aged men

Iqbal A. R. Al-Shayji, Muriel J. Caslake, Jason M. R. Gill


Prior moderate exercise reduces plasma triglyceride (TG)-rich lipoprotein concentrations, mainly in the large very low-density lipoprotein (VLDL1) fraction, but the mechanism responsible is unclear. We investigated the effects of brisk walking on TG-rich lipoprotein kinetics using a novel method. Twelve overweight/obese middle-aged men underwent two kinetic studies, involving infusion of Intralipid to block VLDL1 catabolism, in random order. On the afternoon prior to infusion, subjects either walked on a treadmill for 2 h at ∼50% maximal oxygen uptake or performed no exercise. Multiple blood samples were taken during and after infusion for separation of Intralipid (Sf 400) and VLDL1 (Sf 60–400). VLDL1-TG and -apoB production rates were calculated from their linear rises during infusion; fractional catabolic rates (FCR) were calculated by dividing linear rises by fasting concentrations. Intralipid-TG FCR was determined from the postinfusion exponential decay. Exercise reduced fasting VLDL1-TG concentration by 30% (P = 0.007) and increased TG enrichment of VLDL1 particles [30% decrease in cholesteryl ester (CE)/TG ratio (P = 0.007); 26% increase in TG/apoB ratio (P = 0.059)]. Exercise also increased VLDL1-TG, VLDL1-apoB, and Intralipid-TG FCRs by 82, 146, and 43%, respectively (all P < 0.05), but had no significant effect on VLDL1-TG or -apoB production rates. The exercise-induced increase in VLDL1-apoB FCR correlated strongly with the exercise-induced changes in VLDL1 CE/TG (r = −0.659, r = 0.020) and TG/apoB (r = 0.785, P = 0.002) ratios. Thus, exercise-induced reductions in VLDL1 concentrations are mediated by increased catabolism, rather than reduced production, which may be facilitated by compositional changes to VLDL1 particles that increase their affinity for clearance from the circulation.

  • triglyceride
  • lipoprotein
  • very low-density lipoprotein
  • fractional catabolic rate

elevated plasma concentrations of triglyceride (TG)-rich lipoproteins are associated with increased risk of vascular disease and are implicated in the atherosclerotic disease process by a number of direct and indirect mechanisms (24). There is clear evidence that moderate exercise can lower fasting and postprandial TG concentrations by 20–25% in population groups at increased risk of cardiovascular disease, such as centrally obese middle-aged men (16, 18, 23) and postmenopausal women (19). This is largely an acute effect elicited by recent exercise rather than a chronic training adaptation (20).

The exercise-induced TG reduction is quantitatively larger in hepatically derived VLDL (16, 18), particularly in the large VLDL1 (Sf 60–400) fraction (16), than in intestinally derived chylomicrons. As high concentrations of VLDL1 are the major determinant of elevated plasma TG levels (27, 39), and VLDL1 are the primary precursor particles for atherogenic small-dense LDL (34), reducing VLDL1 concentration is likely to induce clinically important changes to the atherosclerotic risk profile. Exercise-induced reduction in circulating VLDL1 could reflect reduced hepatic VLDL1 production, increased lipoprotein lipase (LPL)-mediated VLDL1 clearance, or a combination of the two. Stable-isotope kinetic studies have demonstrated that, in nonobese, recreationally active young men, moderate-intensity exercise sessions of 90–120 min duration can increase clearance of total VLDL-TG (29, 40) and decrease total hepatic VLDL-apolipoprotein B (apoB) production (29). However, these studies considered all VLDL as a single lipoprotein class, and lipoprotein kinetic studies have shown that VLDL is metabolically heterogeneous, with accumulating evidence demonstrating that both the production and catabolism of large TG-rich VLDL1 (Sf 60–400) and smaller cholesterol-rich VLDL2 (Sf 20–60) are regulated independently of each other (17, 35). Thus, effects of exercise on kinetics of VLDL1, the class of lipoprotein most affected by exercise (16), may not be fully revealed in studies that consider VLDL as a single entity. Furthermore, studies investigating the effects of moderate exercise on TG-rich lipoprotein kinetics in overweight/obese middle-aged men, a group in which the atherogenic lipoprotein phenotype (i.e., elevated TG, low HDL, and small dense LDL) is common (5, 31), and a typical population at which exercise-for-health guidelines are targeted, are lacking. This is important, as moderate exercise may have potential to be used as a first-line therapeutic option for preventing and treating the primary defect in obesity/insulin resistance-related dyslipidemia.

Therefore, the aim of the present study was to investigate the effects of a session of brisk walking on TG-rich lipoprotein (TRL) kinetics [i.e., VLDL1-TG and -apoB production and fractional catabolic rates (FCR) and Intralipid-TG FCR (as a surrogate measure of chylomicron-TG FCR)] in a group of overweight/obese middle-aged men. This was conducted using a recently established novel method to determine TRL kinetics using an infusion of Intralipid (3), which relies on the fact that Intralipid, due to its similarity to chylomicrons (34), competes with VLDL1 for the same LPL-mediated clearance pathway (6).



Twelve overweight/obese men, age 44.0 ± 8.4 yr [mean ± SD], body mass 95.9 ± 17.1 kg, BMI 31.3 ± 5.1 kg/m2, waist circumference 106.9 ± 13.2 cm, and maximal oxygen uptake (V̇o2max) 37.9 ± 7.2 ml·kg−1·min−1 participated in this study. All subjects were apparently healthy, normotensive, normoglycemic, non-smokers who displayed no symptoms of coronary heart disease during a clinical exercise stress test. None was taking any drugs known to affect lipid or carbohydrate metabolism, and none had the E2/2 phenotype. The study was conducted with the approval of the North Glasgow University Hospitals NHS Trust Ethics Committee, and subjects gave written informed consent prior to participation.

Study design.

Each subject underwent two kinetic studies within an interval of 2 wk. Preconditions (i.e., control and exercise) were different and administered in random order. On the afternoon prior to the study day, subjects either walked 120 min on a treadmill at an intensity of ∼50% V̇o2max (Exercise trial) or performed no exercise (Control trial). They were asked to weigh and record their dietary intake for 2 days prior to their first study day and replicate this for the second study and were instructed to refrain from alcohol consumption on those 2 days. No exercise other than the treadmill walk in the Exercise trial was performed during the three days prior to each study day.

Treadmill walk.

A preliminary submaximal incremental treadmill test (1) was performed at least 1 wk prior to the first study day to estimate the gradient necessary to elicit an intensity corresponding to 50% of their V̇o2max. The 120-min walk in the Exercise trial was completed ∼16–18 h before the Exercise kinetic study; this time interval was chosen as it reflects the time scale at which the TG-lowering effects of exercise appear to be maximized (20). At 15-min intervals during the walk, expired air samples were collected using Douglas bags for the determination of oxygen uptake and carbon dioxide production, heart rate was recorded by short-range telemetry (Polar S610i, Polar Electro, Finland), and ratings of perceived exertion (8) were obtained.

Kinetic study.

TRL kinetics (i.e., VLDL1-TG and -apoB production and FCR as well as Intralipid-TG FCR) were determined using the “Intralipid Method” as previously described (3). Briefly, subjects reported to our Clinical Investigation Suite on the morning of the kinetic study day after a 12-h fast. They were given a bolus dose of 0.1 g/kg of 20% Intralipid followed immediately by an intravenous infusion of 0.1 g·kg−1·h−1 of 10% Intralipid for 75 min. Multiple EDTA blood samples were drawn at baseline, at 10-min intervals during infusion, and at 2.5, 5, 10, 15, 20, 30, 40, and 60 min postinfusion. During the infusion, Intralipid is cleared by LPL in preference to VLDL1, with >90% of VLDL1 catabolism being blocked by the presence of Intralipid (7). This leads to an accumulation of VLDL1 in the circulation in proportion to its rate of hepatic production (3). Accordingly, VLDL1-TG and -apoB production rates were calculated from the linear increases in the circulating VLDL1 fraction during infusion, as described previously (3). Intralipid-TG FCR was determined from its exponential decay after infusion (3). The FCRs of VLDL1-TG and -apoB were calculated from the gradient of the linear increase in their concentrations (in mg/dl) over time (min) divided by fasting concentrations (in mg/dl) and then multiplied by 60 min and 24 h (3).

Samples for lipoprotein separation were stored at 4°C until processing within 24 h of collection. At each time point, Intralipid (Sf >400) and VLDL1 (Sf 60–400) fractions were separated from whole plasma by density gradient ultracentrifugation as previously described (3). TG concentrations were measured in the Intralipid and VLDL1 fractions at all time points and were corrected for glycerol (3) by use of kits described below. It is not possible to determine VLDL2 kinetics using the Intralipid method, as VLDL2 are not solely cleared from the circulation by LPL; thus, Intralipid infusion does not completely block their clearance.

Analytic methods.

apoB concentrations were measured directly in the VLDL1 fraction by automated immunoturbidimetry using a commercially available kit (Wako Chemicals, Neuss, Germany). In four subjects, where VLDL1-apoB was not detectable by the immunoturbidimetric method in at least one of their trials, apoB concentrations were measured manually by precipitation with isopropanol (11) with subsequent protein measurement using a modified Lowry method (28) for both Exercise and Control trials.

Plasma glucose (Randox Laboratories, Crumlin, UK) and total and HDL cholesterol (Roche Diagnostics, Mannheim, Germany) were analyzed in the fasted state using commercially available kits. VLDL1-apoC-II, -apoC-III, and -apoE were also measured in the fasted state using commercial automated turbidimetric immunoassay kits (Wako Chemicals). LDL cholesterol was calculated in the fasted state using the Friedewald equation (15). TG (Roche Diagnostics), NEFA (Wako Chemicals), and glycerol (Randox Laboratories) were analyzed at all time points. In the VLDL1 fraction, total protein was measured using a modified Lowry assay (28), and free cholesterol (FC) and phospholipids (PL) were measured using commercially available kits (Wako Chemicals). VLDL1 concentration was calculated by summing the concentrations of its components in milligrams per deciliter [i.e., TG, FC, cholesteryl ester (CE), PL and protein], where CE concentration was determined by multiplying the difference between total cholesterol and FC concentrations (in mg/dl) by 1.68 to account for the difference in mass between cholesterol and CE.

apoE phenotype was determined for each subject by isoelectric focusing using Western blot techniques as described by Menzel (33) and Havekes (26). Insulin was measured in EDTA plasma using commercially available ELISA kits (Mercodia, Uppsala, Sweden).

Power calculation.

Previous data from five subjects who underwent two Intralipid infusions at the 0.1 and 0.2 g·kg−1·h−1 Intralipid doses indicated that the standard deviations for within-subject variability in VLDL1-apoB and VLDL1-TG production were 12.7 and 20.1%, respectively, when preceding diet, exercise, and alcohol intake were well controlled (3). Given that a prolonged session of moderate exercise, typically reduces fasting and postprandial TG concentrations by 20–25% (an effect detectable with groups of 8–12 subjects) (20), we sought to power the study to detect at least a 20% difference in VLDL-TG production. A sample size of 12 was chosen as this would enable detection of an 11% difference in VLDL1-apoB production and an 18% difference in VLDL1-TG production, with 80% power at the 0.05 level.

Calculations and statistics.

The number of apoC-II, apoC-III, and apoE molecules per VLDL1 particle were calculated, in the fasted state, by dividing the apolipoprotein concentrations (in mmol/l) by the apoB concentration (in mmol/l). In addition, VLDL1 TG/apoB ratio (mol:mol), representing the size of the VLDL1 particle in the circulation; CE/TG ratio (mol:mol), representing the composition of its core; CE/apoB ratio (mol:mol), representing the CE content per VLDL1 particle; and FC/PL ratio (mg:mg), representing the composition of its surface layer, were also calculated in the fasted state.

Net energy expenditure during the 120-min walk was calculated using indirect calorimetry assuming no protein oxidation (14). HOMA-IR (homeostasis model assessment of insulin resistance) was used as a validated surrogate measure of insulin resistance (32).

Statistical analyses were performed using Minitab for Windows (v. 14.0; MINITAB, State College, PA) and Statistica (Release 6.0, StatSoft). Normality of data was checked using the Anderson-Darling test. When data did not approximate a normal distribution, these were log-transformed; namely insulin, HOMA-IR, VLDL1-apoB fasting concentrations, TG/apoB ratios, FC/PL ratio, Intralipid-TG FCR, VLDL1-apoB FCR and production rate, VLDL1-TG production rate, and the exercise-induced change in VLDL1-apoB FCR and production rate required transformation. Comparisons between Control and Exercise trials were made using paired t-tests. Relationships between variables were determined by Pearson product-moment correlations. Significance was accepted at the P < 0.05 level. Data are presented as means ± SE unless otherwise stated.


Treadmill walk.

Subjects walked at a speed of 4.5 ± 0.2 km/h up a gradient of 6.0 ± 0.6%. All subjects completed the 120-min walk without difficulty, with an average perceived exertion of 12.2 ± 0.4 (between “fairly light” and “somewhat hard”) on the Borg scale of 6–20 (8). Mean V̇o2 was 18.6 ± 0.8 ml·kg−1·min−1 (49.2 ± 0.7% V̇o2max), mean heart rate was 123 ± 3 beats/min, and net energy expenditure for the walk was 3.5 ± 0.1 MJ (837 ± 35 kcal).

Fasting concentrations in control and exercise trials.

Mean plasma and VLDL1 composition and concentrations in the fasted state are shown in Table 1. Fasting VLDL1-apoE concentrations were below limits of detection in two subjects in at least one trial; thus, apoE data are presented for n = 10. Exercise significantly reduced fasting plasma TG and glucose concentrations by 21 and 3%, respectively, and increased NEFA by 16% (all P < 0.05). There was a reduction of 10% in insulin concentrations and of 13% in HOMA-IR postexercise, but these did not quite achieve statistical significance (P = 0.080 and 0.059, respectively). Fasting VLDL1 concentrations were significantly lower by 34% after exercise (P = 0.006), as were fasting concentrations of VLDL1-TG (by 31%), VLDL1-cholesterol (by 40%), VLDL1-apoB (by 44%), VLDL1-apoC-II (by 28%), VLDL1-apoC-III (by 27%), and VLDL1-apoE (by 31%, n = 10) (all P <0.05). There were no significant differences in fasting concentrations of total, HDL-cholesterol, and LDL-cholesterol between trials.

View this table:
Table 1.

Plasma and VLDL1 concentrations in the fasted state in Control and Exercise trials

VLDL1 compositional responses to exercise.

Table 2 shows VLDL1 composition in the fasted state in the Control and Exercise trials. There was a tendency for circulating VLDL1 particles to be larger following exercise, with the TG/apoB ratio being 26% greater (P = 0.059) in the Exercise than in the Control trial. In addition, the TG enrichment of the core of the VLDL1 particle was markedly increased following exercise with VLDL1 particles containing 8.3 TG molecules for every CE molecule in the Exercise trial compared with 5.9 TG molecules per CE molecule in the Control trial (P = 0.007). This change in CE/TG ratio was almost entirely the consequence of an increase in TG enrichment of the VLDL1 particle: in contrast to the higher VLDL1 TG/apoB ratio in the Exercise trial, the VLDL1 CE/apoB ratio was essentially unchanged by exercise (9% decrease, P = 0.537). PL was not detectable in VLDL1 in one subject; consequently, the FC/PL ratio is presented for 11 subjects. The FC/PL ratio, which reflects the composition of the VLDL1 surface layer, did not differ between the Control and Exercise trials (0% change, P = 0.606). In addition, there were no significant differences between trials in the fasting ratios of apoC-II/apoB, apoC-III/apoB or apoE/apoB, although these ratios were all numerically higher in the Exercise trial than in the Control trial, by 31% (P = 0.23), 31% (P = 0.10), and 35% (P = 0.409), a factor that is likely to reflect the fact that the VLDL1 particles were larger following exercise.

View this table:
Table 2.

VLDL1 particle composition in the fasted state in Control and Exercise trials

TRL kinetics in response to exercise.

Figure 1 shows mean (±SE) concentrations in Intralipid-TG, VLDL1-TG, and VLDL1-apoB over the course of the infusion. Figure 2 shows mean (±SE) Intralipid-TG concentrations postinfusion. VLDL1-TG and -apoB production rates and FCRs and Intralipid-TG FCRs calculated, on an individual basis, from the linear rises in VLDL1-TG and VLDL1-apoB concentrations during the infusion, and the exponential fall Intralipid-TG concentrations postinfusion are shown in Table 3. Although exercise did not significantly change VLDL1-TG or VLDL1-apoB production rates, there were slight, nonsignificant increases of 13% (P = 0.104) and 12% (P = 0.338) in VLDL1-TG and VLDL1-apoB production, respectively, after exercise. In contrast, exercise induced marked increases in the VLDL1-TG and VLDL1-apoB FCRs of 82% (P = 0.002) and 146% (P = 0.015). The Intralipid-TG FCR also increased significantly with exercise, although at 43% (P = 0.001) the relative increase was about half as great as the increase in the VLDL1-TG FCR.

Fig. 1.

Intralipid-TG (top), VLDL1-TG (middle), and VLDL1-apoB (bottom) concentrations during Intralipid infusion in Control and Exercise trials. Values are means ± SE; n = 12.

Fig. 2.

Postinfusion Intralipid-TG concentrations in Control and Exercise trials. Values are means ± SE; n = 12.

View this table:
Table 3.

VLDL1-TG, VLDL1-apoB, and Intralipid kinetics in Control and Exercise trials

Correlates with kinetic variables.

A number of expected relationships between variables were evident in both the Control and the Exercise trials. Increasing insulin resistance and adiposity were associated with increasing VLDL1-TG production: there were significant positive correlations between BMI, waist circumference, and HOMA-IR on one hand and VLDL1-TG production on the other in both Control and Exercise trials (r = 0.649 to 0.791, all P < 0.05). Increasing fasting TG concentrations and fasting VLDL1-TG concentrations were associated with decreasing Intralipid-TG and VLDL1-TG FCRs in both trials (r = −0.695 to −0.895, all P < 0.05). In the Exercise trial, fasting TG concentration and fasting VLDL1-TG concentration were associated with VLDL1-TG production (r = 0.661 and r = 0.705, respectively, all P < 0.05).

Of particular interest was the effect of exercise on the relationship between VLDL1-TG FCR and Intralipid-TG FCR. As Intralipid-TG and VLDL1-TG are cleared by the same mechanism (i.e., TG hydrolysis by LPL), a significant relationship between these two variables would be predicted, and indeed, strong significant correlations were observed between these variables in both trials (Control: r = 0.91, P < 0.0005; Exercise: r = 0.82, P = 0.001). However, the nature of the relationship between Intralipid-TG FCR and VLDL1-TG FCR (i.e., the Intralipid-TG FCR/VLDL1-TG FCR ratio) was changed significantly by exercise (Table 3). In the Control trial, one pool of VLDL1-TG was cleared for every 3.17 pools of Intralipid-TG cleared. However, in the Exercise trial, the affinity of VLDL1-TG for clearance relative to Intralipid-TG was increased such that one pool of VLDL1-TG was cleared per 2.48 pools of Intralipid-TG cleared: a 28% increase in the clearance rate of VLDL1-TG relative to Intralipid-TG compared with the Control trial (P = 0.037).

In addition, the VLDL1 TG/apoB ratio in the fasted state was positively correlated with VLDL1-apoB FCR in both Control (r = 0.624, P = 0.030) and Exercise (r = 0.634, P = 0.027) trials; i.e., VLDL1-apoB clearance was faster for larger, more TG-enriched VLDL1 particles. Furthermore, the change in VLDL1 TG/apoB ratio following exercise was strongly correlated with the exercise-induced change in the VLDL1-apoB FCR (r = 0.785, P = 0.002; Fig. 3). Thus, the increased VLDL1-apoB FCR following exercise was strongly associated with the increased size and TG enrichment of the VLDL1 particles.

Fig. 3.

Scatterplots showing relationships between change in VLDL1-apoB FCR (fractional catabolic rate) from Control to Exercise trial and change in VLDL1 TG/apoB ratio from Control to Exercise trial (top) and between change in VLDL1-apoB FCR from Control to Exercise trial and change in VLDL1 CE/TG ratio from Control to Exercise trial (bottom). Linear regression lines and correlation coefficients for the relationships plotted.

In a similar vein, there was a significant negative correlation between the fasting VLDL1 CE/TG ratio and the VLDL1-TG FCR in the Control trial (r = −0.727, P = 0.007). This relationship was borderline significant in the Exercise trial (r = −0.541, P = 0.070). In other words, less TG enrichment of the VLDL1 core was associated with slower VLDL1-TG clearance. The exercise-induced change in VLDL1 CE/TG ratio was significantly negatively associated with exercise-induced change in the VLDL1-apoB FCR (r = −0.659, P = 0.020; Fig. 3) and had a borderline negative association with the exercise-induced change in VLDL1-TG FCR (r = −0.504, P = 0.095).


The main findings of this study were 1) that 120 min of brisk walking significantly increased the FCRs of VLDL1-TG and -apoB and of Intralipid-TG but did not significantly affect VLDL1-TG or -apoB production in middle-aged overweight/obese men; 2) that exercise increased the VLDL1-TG FCR to a greater extent than the Intralipid-TG FCR; 3) that exercise increased TG enrichment of VLDL1 particles (as assessed by the CE/TG ratio) and tended to increase VLDL1 particle size (as assessed by the TG/apoB ratio); and 4) that the exercise-induced increase in VLDL1-apoB FCR was strongly correlated with the exercise-induced change in VLDL1 size and the exercise-induced change in VLDL1 TG-enrichment.

The present findings are broadly consistent with two earlier stable-isotope studies investigating the effect of moderate exercise on total VLDL kinetics in young nonobese men which also reported that exercise increased total VLDL-TG clearance rather than decreased VLDL-TG production (29, 40). Our data expand on these earlier findings in a number of important ways. First, we studied middle-aged overweight and obese men, who are a particularly clinically relevant group to study as a potential target group for intervention. The present subjects had fasting plasma TG concentrations (∼1.5 mmol/l) in the range associated with increased cardiovascular disease risk (9) and possessed an atherogenic lipoprotein phenotype (25), so this alteration to TG-rich lipoprotein metabolism, which led to a 21% lowering of plasma TG, is likely to be of clinical relevance. Second, the novel methodological approach utilized in the present study, which involved an infusion of Intralipid, uniquely enabled simultaneous measurement of both VLDL1-TG and chylomicron-like Intralipid-TG metabolism in response to a prior session of moderate exercise. Third, we considered the metabolism of large TG-rich VLDL1, rather than of total VLDL, as a single entity, which is important because it is becoming increasingly clear that large VLDL1 and smaller VLDL2 are regulated independently of each other (17, 35); so examining the kinetics of individual VLDL subfractions provides a more accurate model of in vivo lipoprotein metabolism than studies of total VLDL kinetics can provide (16).

Intralipid is a “chylomicron-like” artificial lipid emulsion, whose particles are metabolized by the action of LPL, with similar kinetics to intestinally derived chylomicrons (34). As such, Intralipid-TG kinetics can be viewed as providing a surrogate measure of chylomicron-TG kinetics. The literature on exercise and chylomicron-like particle TG kinetics suggests that a threshold volume of exercise is needed to increase chylomicron-like TG clearance. Whereas prolonged (≥3 h) exercise has been shown to increase chylomicron-like TG clearance (4, 10, 37), 90 min of brisk walking did not increase chylomicron-like TG clearance, despite reducing TG concentrations, in middle-aged men (21). As increased chylomicron-like TG clearance is largely the consequence of increased LPL activity, the relationship between chylomicron-like-TG clearance and post-heparin LPL activity is very strong (12, 36), the latter observation (i.e., lower TG concentrations despite similar chylomicron-like TG clearance) (21) suggests that mechanisms other than increased LPL activity are likely to also contribute to the TG-lowering effect of exercise. Indeed, a number of studies have demonstrated a dissociation between TG lowering and increased LPL activity following exercise (22). In the present study, the 43% increase in Intralipid-TG FCR in response to 2 h of walking is likely to reflect increased LPL activity, although exercise-induced changes in blood perfusion of LPL-rich tissues resulting in increased interactions between Intralipid-TG and LPL could conceivably have also contributed (30).

However, increases in LPL activity (or increases in blood perfusion) alone cannot explain why exercise increased the VLDL1-TG FCR to a greater extent than the Intralipid-TG FCR. Although chylomicron-TG (or Intralipid-TG) and VLDL1-TG are cleared by the same LPL-mediated pathway, the affinity of the larger chylomicron/Intralipid particles for LPL-mediated clearance is greater than that of the relatively smaller VLDL1 particles (6). The present data suggest that exercise significantly increased the affinity of VLDL1 particles for clearance relative to that of chylomicron-like Intralipid particles by 28% (Intralipid-TG FCR/VLDL1-TG FCR ratio decreased from 3.17 in the Control trial to 2.48 in the Exercise trial) and that the exercise-induced increase in VLDL1-TG FCR was twice as large as the increase in Intralipid-TG FCR (82 vs. 43%). This helps to explain earlier observations that prior moderate exercise reduces concentrations of VLDL1-TG to a greater extent than chylomicron-TG (16), revealing that the mechanism underpinning this differential effect of exercise on chylomicron-TG and VLDL1-TG concentrations is exercise upregulating clearance of VLDL1-TG to a greater extent than clearance of TG in chylomicron(-like) particles, rather than the alternatively hypothesized mechanism that acute moderate exercise induces a decrease in hepatic VLDL1-TG production (20).

In the present study, exercise induced compositional changes to the VLDL1 particle: VLDL1 particles in the circulation were ∼26% bigger following exercise (P = 0.059) and more TG-enriched VLDL1 particles had ∼8 TG molecules per CE molecule in their core postexercise compared with ∼6 TG molecules per CE molecule in the Control trial (P = 0.007). This is consistent with previous reports that TG enrichment of total VLDL (29) and VLDL1 (16) is increased by prior exercise. However, the present data extend these earlier findings by suggesting that the increased size and TG enrichment of VLDL1 particles following exercise contributed to their enhanced clearance from the circulation. It is well known that larger, more TG-enriched particles are more favorable substrates for LPL-mediated lipolysis (13), and the TG/apoB ratio of VLDL particles correlates strongly with their rate of TG hydrolysis by LPL in an in vitro system (38). In line with this, the present findings indicate that both the exercise-induced changes in VLDL1 particle size (TG/apoB ratio) and TG enrichment (CE/TG ratio) were strongly correlated to the exercise-induced change in the VLDL1-apoB FCR, explaining 62 and 43% of its variance, respectively. However, further study is needed to determine why circulating VLDL1 particles are larger and more TG enriched following exercise and also to verify whether this change in VLDL1 composition was the causal factor leading to the increased affinity of VLDL1 for clearance following exercise.

A particular point of interest is that, in contrast to the agreement between the findings of the two studies on VLDL1-TG kinetics, our findings with respect to VLDL1-apoB kinetics appear to contrast with the earlier observations of Magkos et al. (29), who reported a significant reduction in total VLDL-apoB production rate with no significant change in VLDL-apoB FTR. This may well reflect differences in the nature of the VLDL fraction considered in the two studies: large VLDL1 (Sf 60–400) in the present study vs. total VLDL (Sf 20–400) in the earlier report. In a normolipidemic individual, approximately three-quarters of total VLDL-TG but only about one-third of VLDL-apoB is contained in the VLDL1 fraction (2). Thus, if VLDL are considered in a single lipoprotein pool, overall VLDL-apoB kinetics will be dominated by the kinetics of VLDL2 (Sf 20–60), whereas overall VLDL-TG kinetics will be dominated by the kinetics of VLDL1. Thus, it seems likely that the lower VLDL-apoB100 production rate reported by Magkos et al. would reflect, to a large extent, reduced production of VLDL2. This interpretation is supported by the fact that the VLDL-TG production rate in that study was unchanged by exercise. This line of reasoning can also explain why an effect on VLDL-apoB FTR was not seen by Magkos et al., in contrast to the nearly 150% increase in VLDL1-apoB FCR observed in the present study: transfer of a VLDL1 particle to VLDL2 would not be detected in a model that considers all VLDL in a single pool. These observations reinforce the added value of considering VLDL1 separately from VLDL2 in studies of human lipoprotein kinetics. However, one limitation of the Intralipid method is that it only enables determination of chylomicron-like particle and VLDL1 kinetics; so further investigation would be necessary to verify whether exercise has a direct effect on production of VLDL2 particles.

In summary, the present findings provide insight into the mechanisms by which moderate exercise influences TRL metabolism in overweight/obese middle-aged men. Exercise induced clinically relevant reductions to plasma TG concentrations, significantly increasing clearance of VLDL1-TG and -apoB, and of Intralipid-TG. The data suggest that exercise increased the affinity of VLDL1 for clearance from the circulation, an effect that was related to the exercise-induced increase in VLDL1 particle size and TG enrichment. However, further study is needed to elucidate how exercise alters VLDL1 size and composition and the precise mechanisms underlying the increased affinity of these postexercise VLDL1 particles for clearance.


This work was supported by the British Heart Foundation (PG/06/122/21613) and the Glasgow Royal Infirmary Endowment Fund (05REFCH08).


No conflicts of interest, financial or otherwise, are declared by the author(s).


Author contributions: I.A.R.A.-S., M.C., and J.M.R.G. conception and design of research; I.A.R.A.-S. and J.M.R.G. performed experiments; I.A.R.A.-S., M.C., and J.M.R.G. analyzed data; I.A.R.A.-S., M.C., and J.M.R.G. interpreted results of experiments; I.A.R.A.-S. and J.M.R.G. prepared figures; I.A.R.A.-S. and J.M.R.G. drafted manuscript; I.A.R.A.-S., M.C., and J.M.R.G. edited and revised manuscript; I.A.R.A.-S., M.C., and J.M.R.G. approved final version of manuscript.


We thank Dr. Lesley Hall and Agnes McGinty for clinical assistance on study days and John Wilson, Heather Collin, Josephine Cooney, and Elizabeth Murray for technical assistance.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
View Abstract