Atherosclerosis has been proposed as a phenomenon that is largely influenced by metabolic conditions generated by the postprandial state (35). It has been recently confirmed from epidemiological studies that elevation of postprandial plasma triglyceride concentrations (PPTG) better predicts cardiovascular events such as myocardial infarction than fasting plasma triglyceride (TG) (1,22). Exercise is one of the powerful ways to combat elevation of PPTG and thus may reduce the risk of atherosclerosis (8,11–13,18). Because the effect of exercise on PPTG is known to be transient (<60 h) (11), individuals may need to exercise on a regular basis to gain full benefit.
Although the TG-attenuating effect of exercise is promising, it remains unclear whether the TG-attenuating effect of exercise is derived from exercise per se (e.g., exercise intensity effect) or simply from energy deficit. In this regard, Cureton’s group reported their quantitative meta-analyses on PPTG initially in 2003 (24) and subsequently in 2014 (6), suggesting that an exercise-induced energy deficit is the main determinant for improvement in PPTG. In their updated meta-analysis in 2014 (6), exercise intensity was not analyzed or discussed probably because a very limited number of studies on the effect of exercise intensity on PPTG has been conducted with mixed results (14,19,30). From these, it is apparent that, to test the effect of exercise intensity appropriately, several requirements with respect to energy balance should be fulfilled. First, exercise at different intensities must be isoenergetic (i.e., same energy expenditure) above resting metabolic rate (RMR). Second, the diet should be matched in terms of macronutrient composition and total energy intake. Lastly, nonexercise physical activities such as sitting, standing, and walking should be controlled. In terms of the last issue, previous studies on the effect of exercise on PPTG typically controlled for strenuous exercise but not for “nonexercise” activities/body positions (14,19,30). This might be a source of the discrepancy. A recently published study from our laboratory in which subjects were controlled for steps of preceding days demonstrated an intensity effect of exercise by showing that high-intensity intermittent cycle exercise (90% V˙O2peak) reduced PPTG as compared with that in isoenergetic moderate-intensity exercise (50% V˙O2peak) (28). However, there are no studies on the effect of exercise intensity from 25% to 65% V˙O2max on PPTG while controlling for diet and the nonexercise physical activities described previously.
Therefore, in the present study, we determined the efficacy of isoenergetic exercise at two intensities (LOW, approximately 25% V˙O2max, vs MOD, 65% V˙O2max) compared with a sedentary (i.e., sitting) control condition (CON, approximately 90% of intervention period being seated), with careful control for nonexercise physical activity and diet.
Nine healthy, recreationally active young men participated in the present study (age, 24.0 ± 4.0 yr; height, 172.7 ± 6.8 cm; V˙O2max, 51.6 ± 6.3 mL·kg−1·min−1). Participants were allowed to take part in the present study only if they were nonsmokers, normotensive (<140/90 mm Hg), had a body mass index of <30 kg·m−2, have no clinical history of cardiopulmonary and metabolic diseases, and were taking no medication that could affect lipid or CHO metabolism. All participants were fully informed of any possible risks and procedures, with a written informed consent obtained before participation. The study was conducted with approval from the University of Texas at Austin institutional review board.
Research participants reported to the Human Performance Laboratory after a 13-h overnight fast approximately 1 wk before the initiation of the first experimental trial. Upon arrival, body weight and height were measured after voiding. Participants then rested for 30 min before the measurement of RMR while seated. This was immediately followed by a four-stage submaximal treadmill test to determine the velocity versus V˙O2 relation. After 15 min of rest, maximal oxygen consumption (V˙O2max) was determined on a treadmill (Quinton Instrument, Seattle, WA) using a customized protocol based on the submaximal data and each participant’s age-predicted HRmax. The V˙O2max and submaximal linear regression of running speed versus V˙O2 was used to determine running speed to elicit 65% V˙O2max. Oxygen consumption and carbon dioxide production were monitored during both submaximal and maximal exercise protocols using the oxygen and carbon dioxide analyzers (models S-3A/I and CD-3A; AEI Technologies, Inc., Pittsburgh, PA). For the intermittent walking session (LOW), each participant was asked to self-select his walking speed on the treadmill, similar to their daily walking speed to mimic to a certain extent their free-living ambulatory activity.
Each participant performed three intervention trials in a randomized crossover experimental design, each occurring over 4 d with at least 1 wk between trials (Fig. 1). Each trial consisted of three phases: stabilization phase on day 1 (D1) and day 2 (D2), intervention phase of either sitting nonexercise control (CON), prolonged sitting with subsequent 1-h running at 65% V˙O2max (MOD), or isoenergetic intermittent walking at approximately 25% V˙O2max (LOW) on day 3 (D3), and a high-fat tolerance test (HFTT) on day 4 (D4). Throughout each 4-d trial, participants were asked to refrain from any planned exercise except for the intervention exercise performed in the laboratory on D3. Throughout each 4-d trial, all food was provided and matched on the basis of total energy (kcal), macronutrient content, and consumption timing. In addition, participants consumed a standardized low-fat meal the evening before the HFTT. On D4, participants reported to the laboratory by 0720 hours after a 13-h overnight fast. Body weight was measured after voiding. After 5 min of rest, a catheter was inserted into the antecubital vein and a fasting blood sample was collected 10 min before ingestion of a high-fat meal. Participants were asked to consume the high-fat test meal and then sit quietly while reading, watching movies, and/or working on a computer for the entire duration of the HFTT. Blood samples were collected hourly over 6 h after consumption of the high-fat meal, with expired gas collection being performed for 10 min in a chair at 1, 3, and 5 h postprandial.
All food was selected on the basis of each participant’s preference and provided from the laboratory. Food was picked up by each participant the evening before each day (D1–D3) or delivered upon their request. The time they ate each meal was recorded for the first trial and replicated on the subsequent trials. Dinner on the day before the HFTT was low in fat to reduce the potential effect of carry-over of fat previously ingested on PPTG and was consumed from 1830 to 1900 hours in the laboratory in all trials, which permitted participants to be fasted for the same period before the HFTT on D4 (approximately13 h). Total daily energy intake (TEI) of the food provided was estimated on the basis of body mass and physical activity levels (2):
where PA is physical activity level coefficient and PA of 1 (for sedentary) was used for all treatments. Therefore, TEI estimate was meant to create a negative caloric balance for both exercise trials and a net balance for the control non-exercise condition. For the stabilization period (D1 and D2), the TEI provided approximated that needed for their low level of activity (PA = 1.11). For the intervention day (D3), TEI approximated that needed for their sedentary level of activity. Therefore, it created a net negative energy balance for MOD and LOW but a net energy balance for CON.
Physical posture/activity control
The day before D1, a pedometer (Yamax Digi-Walker SW-200 pedometer; Great Performance Ltd., London, United Kingdom) was attached to the waist belt of participants to monitor daily step counts throughout each trial. To monitor body posture/activity and HR on D3, an activity monitor system was attached to the other side of participant’s waist belt on the evening of D2 until the end of each trial (1400 hours on D4) (IDEEA; MiniSun LLC, Fresno, CA) (34). Briefly, body/limb motion was constantly monitored through different combinations of signals from the five sensors, including goniometers attached to the skin, with collection of motion and posture data including sitting, standing, walking, and lying down: two sensors on the anterior side of the upper legs, two sensors on the inferior side of the feet, and one sensor on the middle portion of the sternum. In addition, three-lead ECG was equipped in the activity monitor for continuous HR monitoring. During the stabilization period (D1 and D2), participants were asked to refrain from any planned exercise but were asked to achieve step numbers between 7000 and 7500 steps per day, which is considered “low level of physical activity” (31). In addition, they were asked to achieve the suggested step numbers by distributing their steps evenly throughout the day. On D3, participants performed the CON, LOW, or MOD intervention over the 9-h period. For the CON trial, participants were asked to remain seated throughout much of the day and to keep the daily step count to less than 2000 steps. For the LOW and MOD trials, participants were also asked to step less than 2000 during the entire time when they were not on the treadmill. Throughout the trials, participants were reminded of these diet and physical activity requirements by frequent text messages, e-mails, and/or phone communication.
On D3, participants performed 1-h moderate-intensity running for MOD or energy-matched low-intensity intermittent walking for LOW on a treadmill to achieve the same energy expenditure above their RMR, equivalent to the energy predicted to be used during 1-h running at 65% V˙O2max. For the MOD trial, participants ran on a treadmill for 1 h at 65% V˙O2max from 1700 to 1800 hours. For the LOW trial, participants performed 9 intermittent walking sessions on the treadmill in the laboratory hourly from 0900 to 1800 hours. Intermittent type of walking was adopted in the present study to simulate the intermittent nature of free-living ambulatory activity to some extent and frequently interrupt sitting to avoid a possible negative effect of prolonged sitting. Participants selected their own walking speed on the treadmill at which speed they felt they walk in their free-living condition. Total duration of intermittent walking was calculated by dividing total oxygen consumption that was estimated to be used during the 1-h moderate intensity exercise at 65% V˙O2max above RMR by oxygen consumption per minute during walking at the self-selected speed above RMR. By doing so, total oxygen consumption during exercise above RMR was matched between MOD and LOW. Durations for the first and last sessions of walking were fixed, namely, 30 min for the first and 60 min for the last walking session, respectively. Durations of the middle seven walking sessions were variable to attain the target energy expenditure and averaged 17.8 ± 4.0 min. During the MOD trial, expired gas measurements were performed intermittently (3 × 10 min), whereas it was measured during every other walking session (1, 3, 5, 7, and 9 sessions) during the LOW trial. Energy expenditure and substrate oxidation were estimated using indirect calorimetry (17).
Participants arrived at the laboratory via automobile transportation by 0720 hours in the morning of each HFTT after a 13-h overnight fast. Participants consumed each high-fat test meal at the same time of the day in a period no longer than 5 min. The test meal was provided on the basis of body mass at 16.1 kcal·kg−1 body weight (1119 ± 135 kcal, 1.2 g·kg−1 fat (67.2%), 1.1 g·kg−1 CHO (27.4%), and 0.22 g·kg−1 protein (5.5%), respectively). One hour before the initiation of HFTT, which started at 0800 hours, 0.5 L of plain water was provided to each participant. The amount of water that was consumed ad libitum during HFTT was recorded for replication for the subsequent trials.
Postprandial substrate oxidation and energy expenditure
On D4, postprandial substrate oxidation and energy expenditure were estimated using indirect calorimetry method (17) for 10 min after 10 min of rest in a seated position. Expired air was simultaneously analyzed using oxygen and carbon dioxide analyzers (Models S-3A/I and CD-3A; AEI Technologies, Inc., Pittsburgh, PA), and the volume of expired air, collected into a meteorological balloon, was later determined through a calibrated gas meter (Parkinson-Cowan CD-4; Instrumentation Associates, Inc., New York, NY). These measurements were performed 1, 3, and 5 h after the test meal intake.
All blood samples collected were immediately transferred to K2 EDTA collection tubes (BD Vacutainer; BD, Franklin Lakes, NJ), centrifuged at 2000g for 15 min at 4°C. Plasma was then stored in separate aliquots at −80°C until later analysis. All measurements for each participant were performed in duplicate within the same run. Plasma TG and glucose concentrations were measured by a spectrophotometric method using commercially available kits (Pointe Scientific, Inc., Canton, MI). Plasma free fatty acid (FFA) concentration was measured with a colorimetric assay, as previously described (32). Plasma insulin concentration was measured with a commercially available human insulin enzyme-linked immunosorbent assay (Alpco Diagnostics, Salem, NH). Intraassay coefficients of variation for TG, FFA, glucose, and insulin were 2.26%, 1.17%, 0.99%, and 5.21%, respectively.
Incremental area under the curves (AUC) above fasting concentrations (TG AUCI), an index of postprandial response, for plasma TG responses were calculated using the trapezium rule. Total AUC (TG AUCT) was also calculated. TG AUCT, TG AUCI, fasting plasma concentrations, daily steps, and body postures/activities across trials were analyzed using one-way ANOVA with repeated measures. Variables during exercise at MOD and LOW were analyzed using a paired t-test. For time course responses of plasma concentrations, a two-way ANOVA with repeated measures was performed to examine the treatment effect and the treatment–time interactions. The Fisher least significant difference post hoc test was used to identify differences among times and/or treatments if statistical significance exists. The Pearson product-moment correlation coefficient analyses were performed among TG AUCT, TG AUCI, and whole-body fat oxidation. All data were analyzed using the PASW Statistics Package software version 18.0 for Mac (SPSS Inc., Chicago, IL) and were presented as mean ± SD, unless otherwise indicated. Statistical significance was declared when P value was less than 0.05.
Physiological and Metabolic Responses to Exercise
Physiological and metabolic responses during the 1-h moderate-intensity continuous running and the nine-session intermittent walking interventions are listed in Table 1. In both MOD and LOW trials, total energy expenditures were similar because the exercise interventions were designed to be isoenergetic. Average speed of the intermittent walking was 2.88 ± 0.19 mph. Whole-body CHO oxidation was significantly higher (P < 0.001) and whole-body fat oxidation was significantly lower in MOD compared with those in LOW (P < 0.001). RER was significantly higher in MOD than that in LOW (P < 0.001). Mean HR (bts/min) during treadmill exercise were 164 ± 9 and 97 ± 7 for MOD and LOW, respectively (P < 0.001).
Diet and Physical Activity Control
Daily calorie intake for D1, D2, and D3 were 2683 ± 161 kcal, 2688 ± 174 kcal, and 2521 ± 158 kcal, with 50.9% ± 0.9% as CHO, 30.0% ± 0.7% as fat, and 19.2% ± 0.6% as protein on average, with no difference across trials. As stated in the Methods, to reduce the effect of carry-over of fat previously ingested on PPTG, participants consumed a low-fat meal the evening before HFTT (total calories, 743 ± 85 kcal; percent from fat and CHO, 24.9% ± 1.7% and 55.0% ± 5.0%, respectively).
With respect to the daily step, there was no difference in step counts across trials (6965 ± 644 and 7197 ± 32, 6942 ± 521 and 7091 ± 389, and 7049 ± 402 and 7069 ± 363 for MOD, LOW, and CON, respectively) during the stabilization period (D1 and D2). During the intervention day (D3), they stepped 11,919 ± 563, 25,682 ± 3699, and 1569 ± 259 for MOD, LOW, and CON, respectively (for all, P < 0.001). On D4 until completion of the HFTT (1400 h), step numbers were not different across trials: 542 ± 203, 515 ± 187, and 565 ± 33 for MOD, LOW, and CON, respectively. Furthermore, when excluding steps counted during the exercise intervention (MOD and LOW), step counts in all trials were not different across trials (1754 ± 326, 1669 ± 374, and 1569 ± 259 for MOD, LOW, and CON, respectively).
Body posture and activity
Body posture/activity was monitored from 0900 to 1800 hours on D3 (Fig. 2). According to the 9-h motion/posture analysis (n = 8 because of data loss from one participant), sitting time was 404 ± 33, 226 ± 55, and 476 ± 36 min for MOD, LOW, and CON, respectively (for all, P < 0.001). Sitting time during the MOD trial was approximately 178 min (approximately 3 h) longer than that during LOW trial (P < 0.001). In other words, participants spent 75% of their time sitting in MOD, 42% in LOW, and 88% in CON during the 9-h intervention period. Both LOW and, to a lesser extent, MOD spent significantly less time sitting, compared with that in CON (for all, P < 0.001). However, the difference in sitting time between MOD and CON was approximately 76 min, which was mostly accounted for by the 60-min running exercise in MOD. Furthermore, walking time was significantly higher in LOW (213.9 ± 34 min) compared with that in MOD (8.5 ± 6 min) and CON (6.8 ± 3 min) (for all, P < 0.001), with no difference between MOD and CON (Fig. 2). When excluding time spent on the treadmill, walking times in MOD, LOW, and CON (8.5 ± 6, 10 ± 14, and 6.8 ± 3 min) were not different across trials. Furthermore, standing time was not different across trials (67.7 ± 28, 100.5 ± 62, and 57.5± 35 min), and running time (60 min) was only found in MOD.
Postprandial Substrate Oxidation
Postprandial whole-body fat oxidation was higher in MOD compared with that in LOW by 14.4% (P = 0.004) and by 33.8% (P < 0.001) and compared with that in CON (Table 2). LOW was also higher compared with CON by 17% (P = 0.002). Net energy expenditures for the HFTT were not different across trials (Table 2).
In the fasted state
Plasma FFA, glucose, and insulin concentrations were not statistically different across trials for the fasted states (Fig. 3). However, fasting plasma TG concentration (mg·dL−1) was lower in MOD (mean (SEM), 70.6 (7.8); P = 0.009) and tended to be lower in LOW (74.1 (6.4), P = 0.056) compared with those in CON (85.5 (6.3)).
In the postprandial state
Total plasma TG responses (TG AUCT) were significantly lower in both MOD (mean (SEM), 860.6 (75.1) mg·dL−1 per 6 h, P < 0.001) and LOW (949.4 (93.6) mg·dL−1 per 6 h, P < 0.04) as compared with those in CON (1138.1 (103.3) mg·dL−1 per 6 h), with a strong trend for it also being lower in MOD than those in LOW (P = 0.066). The incremental plasma TG responses (TG AUCI), an index of the postprandial response, was significantly lower in MOD as compared with those in LOW by 17.2% (P = 0.023) and with those in CON by 33.6% (P = 0.001), respectively (Fig. 4). TG AUCI was also lower in LOW as compared with that in CON by 19.8% (P = 0.048) (Fig. 4). For plasma glucose and insulin, there were no significant treatment–time interactions (P > 0.05) (Fig. 3). For insulin, there was no significant treatment effect. However, there was a significant treatment effect for plasma glucose (P < 0.05) (Fig. 3). MOD and LOW significantly reduced plasma glucose concentration compared with that in CON (P = 0.007 and P = 0.021, respectively), with that in MOD lower than that in LOW (P = 0.025). For plasma FFA concentrations, there was no significant treatment–time interaction (P > 0.05, Fig. 3). However, there was a significant treatment effect, namely, those in MOD was significantly higher than those in LOW (P = 0.005) and CON (P = 0.04).
There was no significant correlation between TG AUCI and fat oxidation, although there was a trend for an inverse correlation between TG AUCT and fat oxidation (r = −0.33, P = 0.09).
The major findings of the present study include that under carefully controlled conditions with respect to physical activity and diet, 1) both MOD and LOW reduced the incremental AUC for postprandial plasma TG concentration (TG AUCI) compared with that in CON and 2) MOD reduced TG AUCI as compared with that in LOW. These findings confirm that raising exercise intensity from low to moderate intensity (approximately 30%–65% V˙O2max) has a significant effect on lowering the PPTG excursion.
In the quantitative meta-analysis in 2003, Cureton’s group suggested that there is not an independent intensity effect of exercise on PPTG after acute exercise at exercise intensities ranging from 30% to 70% V˙O2max, with an energy expenditure of approximately 400 kcal or greater, as long as energy expenditure is balanced (24). In their most recently updated meta-analysis (6), exercise intensity was not discussed as a potential moderator for PPTG, although it was suggested that high-intensity interval exercise is the most potent type of exercise for reducing PPTG (6); this conclusion does not necessarily imply an independent effect of exercise intensity because the analysis was not performed for the effect of intensity but instead performed for exercise type. Their main conclusion was that the major determinant of PPTG after an exercise bout is energy deficit (6), as previously suggested (24). Only three studies, to the authors’ knowledge, that directly investigated the effect of exercise intensities (<70% V˙O2max) on PPTG while matching energy balance (14,19,30) at the exercise intensities (range, 25%–70% V˙O2max) have been reported with mixed results. For example, Katsanos et al. (14) investigated the effect of intensity of isoenergetic exercise (25% and 65% V˙O2max) in healthy young men and found that PPTG was significantly reduced after an exercise at 65% V˙O2max but not at 25% V˙O2max despite an energy expenditure of >1100 kcal. Furthermore, PPTG was significantly lower in exercise at 65% V˙O2max as compared with that at 25% V˙O2max. However, Tsetsonis and Hardman (30) found that PPTG were significantly reduced after exercise at two different intensities (32% and 63% V˙O2max), with a similar magnitude of reduction. On the other hand, Mestek et al. (19) found an opposite result in men with metabolic syndrome in that PPTG was only significantly reduced after a lower-intensity exercise (35%–45% V˙O2max) but not after an isoenergetic bout of exercise (500 kcal) at 60%–70% V˙O2max. The discrepancy may be at least in part due to study populations with different metabolic status (i.e., metabolic syndrome). The existence of uncertainty over the effect of the intensity of exercise for reducing PPTG is primarily due to 1) a very limited number of studies and 2) possible insufficiency of controls for nonexercise physical activity thermogenesis (NEAT) and nutrient intake that are required for the appropriate investigation of the effect of exercise intensity per se. However, none of these studies carefully controlled for or reported controls for food intake and NEAT.
In the present study with careful controls for nonexercise activities/postures and diet, we found that MOD (Fig. 4) was more effective in attenuating PPTG, as compared with LOW. Our finding is consistent with the finding of Katsanos et al. (14) (25% vs 65% V˙O2max) but is in contrast to that of Tsetsonis et al. (30) (32% vs 63% V˙O2max). In the case of Katsanos et al. (14), one could argue that the favorable effect of the higher intensity may stem from higher energy deficit of the moderate-intensity exercise than that of low-intensity exercise. This is because in their calculation of exercise energy expenditure, they included RMR, and thus, energy expenditure during the low-intensity exercise was overestimated by approximately 185 kcal. Although not stated clearly, it is also the case for Tsetsonis et al. (30) where they found no difference in PPTG between exercises. Thus, the overestimation could not explain the discrepancy. The present study in which this potential confounding factor was controlled confirmed that there exists an exercise intensity effect on PPTG at an intensity ranging from low (25% V˙O2max) to moderate (65% V˙O2max).
The exercise intensity effect in the present study (MOD vs LOW) may be due to several potential issues. First, it may result from differences in type of exercise such as continuous (MOD) and intermittent (LOW) exercise on PPTG. However, several studies have shown that intermittent exercise at intensities in the range of 40%–70% V˙O2max is equally effective as continuous exercise in attenuating PPTG (9,20,21). Second, it may be due to differences in the magnitude of excess postexercise oxygen consumption (EPOC) after MOD and LOW because EPOC is a function of exercise intensity and duration. In this regard, it has been shown that EPOC after even higher-intensity exercise (70% V˙O2max) for 30 min over the 9-h postexercise period was only approximately 7 L, which is equivalent to approximately 32 kcal (16). However, it is unlikely that this potentially small difference (<50 kcal) in EPOC after MOD and LOW mediates the 17.2% difference in PPTG. Third, it is also possible that the potential intensity effect stems from the differences in the timing of meal intake relative to exercise during the intervention day. In MOD, a large exercise-induced energy deficit (622.9 ± 115.8 kcal) occurred in a relatively short period (i.e., 60 min) between 1700 and 1800 hours, 30 min before the dinner meal, followed by approximately 13 h of fasting, whereas in LOW, several small amounts of energy deficit occurred intermittently throughout the 9-h period (621.3 ± 96.9 kcal between 0900 and 1800 hours). Because exercise-induced energy deficit was matched between MOD and LOW and the energy replacement by the dinner meal (742.5 ± 84.9 kcal) was 20% greater than the exercise-induced energy deficit in both LOW and MOD, it is unlikely that the differences in timing of meals created the “intensity effect of exercise” in the present study. However, it might be ideal if the pattern of exercise in MOD was matched to that of LOW (i.e., nine intermittent moderate-intensity exercises) from an experimental standpoint, necessitating future studies on the pattern of the timing of meal intake relative to exercise on PPTG. Fourth, it is suggested that timing and total fat amount of the test meal are important moderators for PPTG (6). As in the cases of those studies mentioned (14,19,30), in the present study, subjects consumed the “high-fat content” test meal (>0.7 g fat·kg−1 body weight) at an effective timing (approximately 13 h postexercise), within which timing of exercise is suggested to be not a significant moderator for PPTG (3,6). Lastly, it was reported that there is a sex difference in PPTG after an exercise bout (6), yet we tested only male subjects. However, sex difference is no longer an important moderator of PPTG when TG response was expressed as AUC TGI (6), which is the main outcome for these studies mentioned previously and in the present study.
Another finding in the present study is that LOW reduced PPTG compared with CON. Compared with the effect of moderate-intensity exercise on PPTG, it is less clear that low-intensity exercise (approximately 40% V˙O2max) lowers PPTG. Tsetsonis et al. (30) performed a series of studies showing that exercise at 31% V˙O2max for 120 min, but not for 90 min (29), significantly reduced PPTG, indicating that the duration of 90 min and, thus, energy expenditure at that intensity may not be sufficient. However, Miyashita et al. (20) found that a continuous exercise at slightly higher intensity (approximately 42% V˙O2max) but for a much shorter duration (30 min) reduced PPTG (P = 0.051) by approximately 30%. Furthermore, Zhang et al. (33) have shown in males with hypertriglyceridemia that an exercise at 40% for 1 h significantly reduced PPTG. Moreover, at even lower-intensity exercise (<30% V˙O2max) as in the present study, there are only two studies reported to the authors’ knowledge and both did not show a significant reduction in PPTG after a bout of low-intensity exercise. For example, in the case of Pfeiffer et al. (25), the null effect of an exercise bout at 26% V˙O2max for 30 min in reducing PPTG may be due to insufficient energy deficit (approximately 100 kcal) whereas in the case of Katsanos et al.(14), it does not seem to be due to lack of energy deficit (approximately 1130 kcal during 25% V˙O2max for approximately 238 min). In the study by Katsanos et al. (14), low-intensity exercise (25% V˙O2max) lasted approximately 4 h and finished 1 h before the initiation of the HFTT, suggesting that time to fully activate muscle lipoprotein lipase (mLPL) (>4 h) (15) may not have been sufficient. This may explain the discrepancy. It is also possible that potential variations in NEAT may dilute more preferentially the effect of lower-intensity exercise than to that of moderate- to higher-intensity exercises on PPTG. Taken together, we found that exercise at low intensity (approximately 25% V˙O2max for 214 min) is effective in reducing PPTG, although it is still not clear if the TG-attenuating effect of LOW is due to energy deficit created by exercise or exercise per se or both from the present study design (8), necessitating the future investigations.
The reduction in PPTG might be accomplished by 1) reduced VLDL production, 2) increased tissue clearance of circulating TG (VLDL- and chylomicron-TG), or both. First, previous exercise may decrease PPTG by attenuating VLDL production and secretion from the liver into the circulation (7). In accordance with this notion, a recent isotope tracer kinetics study (4) demonstrated that previous exercise at 60%–65% V˙O2peak for 120 min significantly reduced plasma TG contents derived endogenously (i.e., VLDL-TG) but not exogenously (i.e., from the meal). This may occur through changes in postprandial energy distribution by previous exercise. For instance, although extramyocellular energy sources (i.e., blood glucose and plasma FFA) are main energy sources for muscle contraction during exercise at 25% V˙O2max, during an exercise at approximately 65% V˙O2max, energy sources are more dependent on intramyocellular sources, mainly muscle glycogen and, to a lesser extent, intramyocellular TG (27). These depleted intramyocellular energy sources must be replenished after exercise from extramyocellular sources, i.e., TG, FFA, and glucose in the blood. Consistent with the notion, a 31P magnetic resonance spectroscopy study showed that previous exercise at 75%–85% of HRmax for 45 min increased net muscle glycogen synthesis more than threefold and decreased de novo lipogenesis and liver TG content after a CHO-rich meal (26). Therefore, the exercise-induced depletion in intramyocellular energy sources, as it occurs during exercise at 65% V˙O2max, may allow more of the postprandial plasma substrates to be taken into the skeletal muscles, a site of TG clearance, for the replenishment of muscle glycogen and intramyocellular triglyceride. This could shunt substrate away from the liver, a site for de novo lipogenesis, leading to less VLDL-TG production and secretion into the circulation (23). In the present study, MOD reduced CHO oxidation and plasma glucose concentrations at the whole-body levels during HFTT compared with those during LOW, suggesting a possibly enhanced muscle glycogen synthesis in MOD. In addition, enhanced fat oxidation that in turn reduces availability of fatty acid for production of VLDL may contribute to the exercise-induced reductions in PPTG. In accordance with this idea, many studies including the present study have found that previous exercise significantly increases whole-body fat oxidation in the postprandial state (4,28), which in turn counteracts the stimulatory effect of increased availability of FFA for VLDL-TG production (4). Furthermore, in our previous study (28), a significant inverse correlation between fat oxidation and TG AUCI was found. However, in the present study, we did not find a significant inverse correlation between fat oxidation and TG AUCI. Importantly, in the present study, we did not measure fat oxidation in the fasted state, when fat oxidation is highest, which might reduce the chance of achieving statistical significance. We did, however, find a trend for an inverse correlation between fat oxidation and TG AUCT (r = −0.33, P = 0.09). Second, previous exercise may increase TG clearance primarily through the activation of mLPL. Because mLPL in fast muscle fibers can be activated to a greater extent with higher-intensity exercise (MOD vs LOW) (10), this may further explain the potency of MOD in reducing PPTG compared with LOW and CON. However, the notion has not been well supported by experimental data. For example, Malkova et al. (18) demonstrated using the arteriovenous balance method that although previous treadmill exercise for 2 h at 60% V˙O2max reduced PPTG, TG extraction across legs was not significantly increased. However, they found that glucose uptake was significantly increased. These findings again signify the possible important role of alterations in endogenous TG production and secretion in attenuating PPTG after an exercise. Lastly, it could not be excluded that exercise may affect the digestion and absorption of dietary fat and/or enterocyte chylomicron-TG metabolism, which ultimately affect the excursion of plasma TG (5).
There are several limitations to the present study. First, the study subjects were only composed of healthy young adult males. Therefore, it is not possible to generalize the conclusion drawn from this present study to the female population and also not to older people. Second, different types of exercise used in the present study (intermittent for LOW vs continuous for MOD) could be another potential limitation. However, as it was discussed previously, it is less likely the case because previous studies have shown that these types of exercise are equally effective in reducing PPTG (9,20,21). Lastly, although we found that the intermittent low-intensity walking was effective in reducing PPTG, it does not seem to be translatable to the general public because the duration of walking in the present study (i.e., approximately 214 min·d−1) is impractical because it is too long for the general public.
In conclusion, running exercise at moderate intensity (65% V˙O2max) was more potent in reducing PPTG as compared with isoenergetic intermittent walking exercise (approximately 25% V˙O2max) under conditions controlled for NEAT and food intake, suggesting a potential unique effect of exercise intensity on PPTG at least in a situation where overall energy balance is negative.
The authors thank the subjects of this study for their conscientious participation.
This study was not funded and served as part of Il-Young Kim’s dissertation.
The authors’ contributions are as follows: I.-Y. K. and E. F. C., conception and design of research; I.-Y. K. and S. P., subject recruitment and data collection; I.-Y. K., S. P., and E. F. C., data analyses; I.-Y. K., S. P., J. R. T, and E. F. C., interpretation of results of the experiments; I.-Y. K., draft of the article. All authors edited and approved the article for submission.
No conflicts of interest, financial or otherwise, are declared by the authors.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. Bansal S, Buring JE, Rifai N, Mora S, Sacks FM, Ridker PM. Fasting compared with nonfasting triglycerides
and risk of cardiovascular events in women. JAMA
. 2007; 298 (3): 309–16.
2. Brooks GA, Butte NF, Rand WM, Flatt J-P, Caballero B. Chronicle of the Institute of Medicine physical activity
recommendation: how a physical activity
recommendation came to be among dietary recommendations. Am J Clin Nutr
. 2004; 79 (5): 921S–30S.
3. Burns SF. Reductions in postprandial lipemia with exercise: is timing important? Med Sci Sports Exerc
. 2008; 40 (7): 1353–4.
4. Davitt PM, Arent SM, Tuazon MA, Golem DL, Henderson GC. Postprandial triglyceride and free fatty acid metabolism in obese women after either endurance or resistance exercise. J Appl Physiol (1985)
. 2013; 114 (12): 1743–54.
5. Duez H, Pavlic M, Lewis GF. Mechanism of intestinal lipoprotein overproduction in insulin resistant humans. Atheroscler Suppl
. 2008; 9 (2): 33–8.
6. Freese EC, Gist NH, Cureton KJ. Effect of prior exercise on postprandial lipemia: an updated quantitative review. J Appl Physiol (1985)
. 2014; 116 (1): 67–75.
7. Gill JM, Frayn KN, Wootton SA, Miller GJ, Hardman AE. Effects of prior moderate exercise on exogenous and endogenous lipid metabolism and plasma factor VII activity. Clin Sci (Lond)
. 2001; 100 (5): 517–27.
8. Gill JM, Hardman AE. Postprandial lipemia: effects of exercise and restriction of energy intake compared. Am J Clin Nutr
. 2000; 71 (2): 465–71.
9. Gill JM, Murphy MH, Hardman AE. Postprandial lipemia: effects of intermittent versus continuous exercise. Med Sci Sports Exerc
. 1998; 30 (10): 1515–20.
10. Hamilton MT, Etienne J, McClure WC, Pavey BS, Holloway AK. Role of local contractile activity and muscle fiber type on LPL regulation during exercise. Am J Physiol
. 1998; 275 (6 Pt 1): E1016–22.
11. Hardman AE, Lawrence JE, Herd SL. Postprandial lipemia in endurance-trained people during a short interruption to training. J Appl Physiol (1985)
. 1998; 84 (6): 1895–901.
12. Herd SL, Kiens B, Boobis LH, Hardman AE. Moderate exercise, postprandial lipemia, and skeletal muscle lipoprotein lipase activity. Metab Clin Exp
. 2001; 50 (7): 756–62.
13. Herd SL, Lawrence JE, Malkova D, Murphy MH, Mastana S, Hardman AE. Postprandial lipemia in young men and women of contrasting training status. J Appl Physiol (1985)
. 2000; 89 (5): 2049–56.
14. Katsanos CS, Grandjean PW, Moffatt RJ. Effects of low and moderate exercise intensity on postprandial lipemia and postheparin plasma lipoprotein lipase activity in physically active men. J Appl Physiol (1985)
. 2004; 96 (1): 181–8.
15. Kiens B, Lithell H, Mikines KJ, Richter EA. Effects of insulin and exercise on muscle lipoprotein lipase activity in man and its relation to insulin action. J Clin Invest
. 1989; 84 (4): 1124–9.
16. Laforgia J, Withers RT, Shipp NJ, Gore CJ. Comparison of energy expenditure elevations after submaximal and supramaximal running. J Appl Physiol (1985)
. 1997; 82 (2): 661–6.
17. Lusk G. Animal calorimetry: twenty-fourth paper. Analysis of the oxidation of mixtures of carbohydrates and fat. J Biol Chem
. 1924; 59 (1): 41–2.
18. Malkova D, Evans RD, Frayn KN, Humphreys SM, Jones PR, Hardman AE. Prior exercise and postprandial substrate extraction across the human leg. Am J Physiol Endocrinol Metab
. 2000; 279 (5): E1020–8.
19. Mestek ML, Plaisance EP, Ratcliff LA, Taylor JK, Wee S-O, Grandjean PW. Aerobic exercise and postprandial lipemia in men with the metabolic syndrome. Med Sci Sports Exerc
. 2008; 40 (12): 2105–11.
20. Miyashita M, Burns SF, Stensel DJ. Accumulating short bouts of brisk walking reduces postprandial plasma triacylglycerol concentrations and resting blood pressure in healthy young men. Am J Clin Nutr
. 2008; 88 (5): 1225–31.
21. Miyashita M, Burns SF, Stensel DJ. Exercise and postprandial lipemia: effect of continuous compared with intermittent activity patterns. Am J Clin Nutr
. 2006; 83 (1): 24–9.
22. Nordestgaard BG, Benn M, Schnohr P, Tybjaerg-Hansen A. Nonfasting triglycerides
and risk of myocardial infarction, ischemic heart disease, and death in men and women. JAMA
. 2007; 298 (3): 299–308.
23. Petersen KF, Dufour S, Savage DB, et al. The role of skeletal muscle insulin resistance in the pathogenesis of the metabolic syndrome. Proc Natl Acad Sci U S A
. 2007; 104 (31): 12587–94.
24. Petitt DS, Cureton KJ. Effects of prior exercise on postprandial lipemia: a quantitative review. Metabolism
. 2003; 52 (4): 418–24.
25. Pfeiffer M, Wenk C, Colombani PC. The influence of 30 minutes of light to moderate intensity cycling on postprandial lipemia. Eur J Cardiovasc Prev Rehabil
. 2006; 13 (3): 363–8.
26. Rabøl R, Petersen KF, Dufour S, Flannery C, Shulman GI. Reversal of muscle insulin resistance with exercise reduces postprandial hepatic de novo lipogenesis in insulin resistant individuals. Proc Natl Acad Sci U S A
. 2011; 108 (33): 13705–9.
27. Romijn JA, Coyle EF, Sidossis LS, Rosenblatt J, Wolfe RR. Substrate metabolism during different exercise intensities in endurance-trained women. J Appl Physiol (1985)
. 2000; 88 (5): 1707–14.
28. Trombold JR, Christmas KM, Machin DR, Kim I-Y, Coyle EF. Acute high-intensity endurance exercise is more effective than moderate-intensity exercise for attenuation of postprandial triglyceride elevation. J Appl Physiol (1985)
. 2013; 114 (6): 792–800.
29. Tsetsonis NV, Hardman AE. Effects of low and moderate intensity treadmill walking on postprandial lipaemia in healthy young adults. Eur J Appl Physiol Occup Physiol
. 1996; 73 (5): 419–26.
30. Tsetsonis NV, Hardman AE. Reduction in postprandial lipemia after walking: influence of exercise intensity. Med Sci Sports Exerc
. 1996; 28 (10): 1235–42.
31. Tudor-Locke C, Bassett DR. How many steps/day are enough? Preliminary pedometer indices for public health. Sports Med
. 2004; 34 (1): 1–8.
32. Zderic TW, Davidson CJ, Schenk S, Byerley LO, Coyle EF. High-fat diet elevates resting intramuscular triglyceride concentration and whole body lipolysis during exercise. Am J Physiol Endocrinol Metab
. 2004; 286 (2): E217–25.
33. Zhang JQ, Ji LL, Fretwell VS, Nunez G. Effect of exercise on postprandial lipemia in men with hypertriglyceridemia. Eur J Appl Physiol
. 2006; 98 (6): 575–82.
34. Zhang K, Werner P, Sun M, Pi-Sunyer FX, Boozer CN. Measurement of human daily physical activity
. Obes Res
. 2003; 11 (1): 33–40.
35. Zilversmit DB. Atherogenesis: a postprandial phenomenon. Circulation
. 1979; 60 (3): 473–85.