| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Division of Cardiology and 2Duke Center for Living, Duke University Medical Center, Durham; and 3Department of Exercise and Sports Science and Human Performance Laboratory, East Carolina University, Greenville, North Carolina
Submitted 20 November 2006 ; accepted in final form 26 March 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
lipoprotein particle size; exercise amount; exercise volume; exercise intensity; physical inactivity
We have reported on the beneficial effects of different amounts and intensities of exercise training on lipoproteins measured within 1 day of the last training session (15). We observed that an exercise amount approximately equal to the caloric expenditure of jogging 17 miles/wk resulted in widespread beneficial effects on lipoproteins and lipoprotein particle size and number. Importantly, we noted that many of the beneficial effects of exercise on lipids and lipoproteins were not observed in the typical lipid profile. Rather, many of the benefits were observed in the effects of exercise on particle size and particle number. This is important as it is now clear that the concentrations of low-density lipoprotein (LDL) particles, small LDL particles, large high-density lipoprotein (HDL) particles, and large very-low-density lipoprotein (VLDL) particles are better indicators of cardiovascular risk than are the elements of the traditional lipid profile (13, 17, 18, 29, 42, 44). We also observed that lower amounts of exercise (calorically equivalent to walking or jogging 
11 miles/wk, regardless of exercise intensity) were clearly and consistently more beneficial for the lipoprotein profile than a continued sedentary lifestyle as observed in an inactive control group.
While the robust, but short-lived, effects of exercise on improving insulin sensitivity are very well described, and importantly, the information has been quite useful for both practical and mechanistic purposes, far less research has been done on the time course of the sustainability of the exercise-induced lipid benefits. Two important issues are whether the beneficial lipid and lipoprotein effects of an aerobic training program last after the cessation of exercise and whether exercise training amount and/or intensity differentially impact the duration of these benefits. These issues are scientifically important in order to be able to mechanistically attribute the observed effects to exercise training rather than to the acute effects from the last session of exercise. They are clinically important by providing information regarding the sustainability of lifestyle effects when a training program is interrupted in the short term. Finally, this information may impact recommendations regarding the optimal and/or minimal exercise prescriptions required for sustained health benefits. Therefore, the main purposes of the present analysis were to determine whether the training-induced benefits in serum lipids and lipoproteins are sustained over 5 and/or 15 days of exercise cessation and to determine if either exercise amount or exercise intensity impact the duration of these benefits.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Subjects. After providing written, informed consent, subjects who were 40 to 65 yr old, sedentary, overweight, or mildly obese (body mass index 25–35 kg/m2) and dyslipidemic (either LDL-cholesterol of 130–190 mg/dl, or HDL-cholesterol <40 mg/dl for men or <45 mg/dl for women) were randomly assigned to one of three training groups or a nonexercising control group. Subjects for this cohort were recruited continuously between January 1999 and June 2002, with exercise training completed by April 2003. Subjects with complete lipid data (n = 240) were used in the analysis.
Exercise training.
The exercise groups were assigned as follows (30) : 1) high-amount/vigorous-intensity exercise, the caloric equivalent of 20 miles of jogging per week for a 90-kg person at 65–80% peak oxygen consumption (
O2); 2) low-amount/vigorous-intensity exercise, the caloric equivalent of 12 miles of jogging per week at 65–80% peak O2; and 3) low-amount/moderate-intensity exercise, the caloric equivalent of 12 miles of walking per week at 40–55% peak O2. For the high-amount/vigorous-intensity group, the specific prescription was to expend 23 kcal·kg body wt–1·wk–1. For the two low-amount groups, energy expenditure was 14 kcal·kg body wt–1·wk–1. For example, if a subject weighed 100 kg, and was randomly assigned to one of the low-amount groups (14 kcal·kg body wt–1·wk–1), then their prescribed weekly caloric expenditure was 14 kcal x 100 kg = 1,400 kcal/wk. The number of minutes of exercise needed per week then depended on their fitness level. If, for example, their peak O2 was 4 liters of oxygen per minute (4 l/min) and they were assigned to the vigorous exercise group, then their exercise work rate was set at a level that required 3 l/min of oxygen consumption (3 l/min is 75% of 4 l/min). The work rate was verified by a submaximal test. The submaximal test involved exercising at increasing work rates until a work rate that required an oxygen consumption of 3 l/min was obtained. The exercise intensity was then maintained at that level by exercising within a heart rate range that corresponded to this initial work rate. The work rate was adjusted at the end of the ramping up period via another maximal and submaximal test. The calculation of caloric expenditure was based on each liter of oxygen consumed during exercise being equivalent to 5 kcal (this was not corrected for respiratory exchange ratio). So, in the above example, an exercise prescription of 1,400 kcal/wk divided by 15 kcal expended/min of exercise at 75% would be obtained by 93 min of exercise per week [(1,400 kcal/wk)/(15 kcal expended/min) = 93 min of exercise/wk].
Details about the prescribed and actual exercise training are included in Table 1. The average weekly exercise frequency for each group is also shown in Table 1. In Table 1 and elsewhere, we describe the amount (defined as kcal of energy expenditure per week) and intensity of exercise as being calorically equivalent to an approximate number of miles of walking or jogging. This was done to help the reader easily understand both the intensity and the amount of exercise prescribed. In fact, very few subjects jogged. Rather, they walked briskly up a moderate to steep grade to achieve the correct intensity (similar to jogging) and/or they exercised on an elliptical trainer or, less often, a stationary bicycle. The low-amount/moderate-intensity and the high-amount groups both averaged 58 min/session for an average of 3.5 or 3.6 times each week, respectively. The low-amount/vigorous-intensity group averaged 43 min per exercise session and 2.9 sessions/wk. Importantly, neither exercise frequency nor exercise duration per session was prescribed but was, within reason, left up to the subject. Rather, the total number of minutes for each subject was prescribed. For example, if a subject was prescribed 200 min/wk, they could choose to do this in four sessions of 50 min each, or five sessions of 40 min each. However, for safety concerns, they were essentially not allowed to choose to do three sessions of 67 min each per week.
|
Crossover controls. In the present study, to minimize the negative effect of randomization to the control group, all participants were assured that if they were in the control group that, after the control period, they would be randomized into one of the three exercise groups. This had the added benefit of increasing the statistical power when exercise detraining analyses were conducted, as in the present study. In previous studies from the STRRIDE trial, the exercise training results were not used from the control crossover subjects as this would violate the independence of groups in statistical procedures comparing the four groups. However, in the present study and for all detraining analyses, all data from subjects that completed the exercise-training period are included in the analysis (for the present report, that included 240 subjects with complete assay data), and the control data are included in the tables and figures for visual comparison.
Dietary evaluations and body weight control. Nutrient intakes of each subject were determined before and after training using a 3-day food record and a 24-h dietary recall interview. To minimize the confounding effects of weight loss, subjects were counseled to maintain diet and body weight. Subjects were recruited with the knowledge that this was an exercise, not a weight loss, study, and as such, if their intent was to lose weight, they should not participate. This was emphasized during the phone screening, during the initial consent meeting, and at the baseline dietary interview. Before written consent was obtained, subjects were informed that if they lost or gained 2.5% or more of their body weight, a nutrition counselor would collect and analyze their recent nutrient intake and suggest methods to maintain or increase their body weight. Body weight was monitored weekly throughout the study.
Lipid and lipoproteins. Fasting plasma samples were analyzed by LipoScience (Raleigh, NC) for lipoprotein profiling by nuclear magnetic resonance spectroscopy (26). Each measurement provides concentrations of six VLDL, four LDL [including intermediate-density lipoprotein (IDL)], and five HDL subclasses and calculated weighted-average VLDL, LDL, and HDL particles sizes, LDL particle concentration, and estimates of total cholesterol, triglycerides, LDL-cholesterol (LDL-C) and HDL-cholesterol (HDL-C). All samples sent for lipid and lipoprotein analysis were anonymous with respect to treatment group.
Statistical methods. Baseline measures are presented as means and SDs. Baseline differences between the four groups were evaluated via ANOVA with Fisher's post hoc test used to determine significant differences between groups. Pre- to posttraining (within 24 h of the last training session) changes in lipoproteins were determined independently for each group by one-tailed t-tests. Five-day and 15-day detraining effects were also evaluated by one-tailed t-tests on the change between baseline samples and samples obtained at each of the detraining time points. One-tailed t-tests were chosen as exercise and lipid reviews suggest that an exercise training intervention would either lead to beneficial effects or no change. We are not aware of data that would suggest that an exercise intervention in sedentary overweight subjects would lead to negative effects on lipids or lipoproteins. Pre- to post control group changes were also evaluated with one-tailed t-tests, in this case with the hypothesis that continued sedentary living would likely lead to either no effects or negative effects on lipids and lipoproteins. Again, we are not aware of any data that would suggest that continued inactivity in overweight sedentary individuals would be expected to lead to improvements in lipids. P values of less than or equal to 0.05 were considered statistically significant. The main purpose of the detraining data analysis was to determine which of the statistically significant acute effects of exercise was still significantly elevated above baseline (i.e., which benefits were sustained) over 5 and/or 15 days of exercise cessation.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
50% more total exercise amount than the two low-amount groups (P < 0.0001). The low-amount/vigorous-intensity group had fewer total exercise minutes per week and a lower weekly exercise frequency than the other two exercise groups (P < 0.0001), as expected. The baseline values for lipids and lipoproteins are presented in Table 2. There were no significant group differences at baseline for any variable with the exception of VLDL size. VLDL size was significantly larger in the moderate-intensity exercise group compared with both the inactive control group and the high-amount exercise group. Outlier analysis revealed four subjects (2 in the moderate-intensity group and 2 in the high-amount group) that were greater than three SDs above the mean. When these subjects were removed from the analyses, no significant group differences remained at baseline. However, as this had essentially no effect on the overall results presented herein, we have chosen to present data from the entire data set, including these subjects.
|
|
|
|
|
O2, visceral fat, and abdominal subcutaneous fat before and after training. Body weight increased in the inactive control group, while decreasing in a dose-response fashion in the exercise groups. Peak O2 decreased in the control group and increased 6.3%, 12.3%, and 17.9% in the low-amount/moderate-intensity, low-amount/vigorous-intensity, and high-amount groups, respectively. The inactive controls experienced a large and significant increase in visceral fat, whereas the two low-amount groups prevented this increase. The high-amount group had a significant decrease in visceral fat. Abdominal subcutaneous fat was unchanged in all but the high-amount group, where it decreased by 6%.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
One of the primary findings in this study was that with respect to high-density lipoprotein particles (HDL-C, large HDL, and HDL size), the beneficial effects found at 24 h were largely sustained at both 5 and 15 days after exercise cessation. This sustained effect was most clearly observed in the high-amount exercise training group where all three variables were significantly improved at all time points after the cessation of exercise training (see Fig. 1). However, this effect was also evident in the two lower-exercise-amount groups, in which increases observed at 24 h were sustained at 5 and 15 days of detraining. Reviews and meta-analyses have previously concluded that regular exercise leads to significant improvements in HDL-C. The present investigation is the first randomized, controlled study to show that this benefit and the improvements in HDL size and large HDL are sustained for up to 2 wk after exercise withdrawal. This benefit may be especially significant clinically, as HDL-C is thought to also have anti-inflammatory, antioxidative, antiaggregatory, anticoagulant, and profibrinolytic activities (25), in addition to its role in reverse cholesterol transport (39).
A second major finding was that the moderate-intensity training group reduced total triglycerides and VLDL-TG at 24 h after exercise training by twice the magnitude as in the two more vigorous exercise-training groups. Perhaps more importantly, only the lower intensity exercise training resulted in sustained total and VLDL-TG-lowering effects after 15 days. In the two vigorous-intensity training groups, total triglycerides and VLDL-TG had returned to baseline after only 5 days, indicating that there was no sustained triglyceride-lowering effect in these two groups. The greater and more prolonged triglyceride-lowering effect that occurred in the lower intensity exercise training group is best explained by the lower exercise intensity itself, as this group had a similar weekly exercise frequency and similar total number of minutes of exercise per week as the high-amount group and equal amount of exercise as the other low-amount exercise group. While the mechanism for this effect is unclear, exercise of different intensities may have tissue-specific effects on the endothelial cell anchored lipoprotein lipase (LPL), resulting in differential effects of exercise of varying intensities on triglyceride, VLDL, and HDL metabolism (12).
Low-intensity exercise equivalent to walking 11 miles/wk appears to reduce triglycerides by 25% in overweight/mildly obese, sedentary, middle-aged men and postmenopausal women (the percentage decrease was the same for normal and hypertriglyceridemic individuals; data not shown). Furthermore, a significant amount of the triglyceride-lowering effect was maintained for up to 15 days after the last session of exercise in the lower intensity group. This represents a highly statistically significant and a clinically important acute and sustained decrease. Although this study is a relatively large, randomized controlled exercise training trial and the findings with regard to triglycerides and low-intensity exercise are important, these findings would be greatly strengthened if replicated by additional large, randomized controlled studies.
As a third major finding, in the present study, we observed that sustained physical inactivity resulted in statistically significant increases in LDL particle number, small dense LDL, and LDL-C and in significant decreases in LDL particle size. In the inactive control cohort, we have previously reported that 6 mo of maintenance of baseline physical inactivity resulted in statistically significant weight gain (36), increased levels of visceral fat (35), increased umbilical waist circumference (36), decreased cardiovascular fitness (4), and decreased insulin sensitivity (11). The increase in LDL particle number and the increase in small dense LDL in the inactive control group observed here are particularly significant indicators of increased cardiovascular risk. Research over the last two decades clearly demonstrates that these variables are better indicators of cardiovascular risk than are those measured in the traditional lipid profile (13, 17, 18, 37, 42). In our previous report on the effects of exercise amounts and intensities on lipoproteins in this cohort, the analysis was designed to look at differences between the exercise groups and the control group and not at changes within a specific group. Thus the observation herein is an important finding that extends the list of the detrimental effects of physical inactivity in humans.
Both the acute and sustained effects of exercise amount and intensity were markedly different with respect to their effects on HDL and VLDL-TG lipoproteins. Generally there is a significant inverse correlation between HDL and triglyceride levels. Indeed, some suggest that decreasing triglycerides play a major role in increasing HDL-C levels(40). However, the lower intensity exercise training was found to have substantial triglyceride-lowering effects that were both acute and sustained. Yet, despite these large effects on triglycerides, no significant changes in HDL-C or large HDL were observed. Similarly, with a higher amount of more vigorous exercise, essentially the opposite occurred with acute and sustained improvements in HDL but only a small acute benefit and no sustained effect on triglycerides. In both of the exercise interventions, the association between changes in triglycerides and HDL was essentially uncoupled as evidenced by a correlation coefficient of 0.003 (P = 0.97). That is to say, the acute and sustained triglyceride effects cannot be attributed or associated with any HDL effect and vice versa. This is a potentially important finding with regard to the dissociation of these two from a mechanistic standpoint.
Hypothesizing about the mechanisms underlying the observed effects of exercise training on lipoprotein levels is challenging, as the interactions between lipid stores, whole body metabolism, and insulin action in muscle, liver, and fat are intimately related and probably all contribute to the observed effects. As our data show, there is a marked discordance in the effects of the exercise training response in VLDL-TGs and HDL metabolism, suggesting that it is likely that these responses have different mechanisms.
Acute changes in lipid levels are likely due to changes in insulin signaling mediated by recent exercise bouts (33). The well-studied rapid reversal of insulin action in muscle following cessation of exercise may be secondary to the rapid change in VLDL metabolism in muscle with enforced inactivity. Specifically, a reduction in VLDL metabolism in muscle with inactivity leads to diversion of fatty acid metabolism away from mitochondrial oxidation and toward increases in triglyceride stores, which has a suppressing effect on insulin action. Changes in HDL-C, however, are harder to achieve, are likely related to changes in body habitus and possibly increases in oxidative capacity in muscle, are likely longer lasting, and are harder to reverse, even with enforced inactivity.
Why moderate-intensity exercise resulted in twice the magnitude of triglyceride lowering as higher intensity training and was the only training program that resulted in sustained decreases in triglycerides despite cessation of training presents an intriguing conundrum. Moderate-intensity exercise relative to high-intensity exercise relies more on lipid as a fuel source. This may imply that moderate-intensity exercise has a longer-lasting effect on lipoprotein lipase in muscle and hepatic lipase in liver. This theoretically leads to increased lipid uptake and oxidation in skeletal muscle, with resultant improvements in insulin action and lower steady-state glucose, insulin, and triglyceride levels (3). Also, high-intensity exercise may lead to a higher rate of triglyceride mobilization from fat stores as a response to the higher catecholamine levels with high-intensity exercise, thereby counteracting any stimuli to reduce VLDL-TG levels through increased utilization and oxidation in muscle. Combined, these effects might explain the relative magnitude of the training-induced lowering of serum VLDL and triglycerides with moderate-intensity exercise relative to vigorous-intensity exercise. Alternatively, vigorous exercise tends to increase mitochondrial capacity relative to lower intensity exercise, and the imbalance between supply and demand created by training cessation and lipid storage may have a detrimental effect on mitochondrial fatty acid oxidation, storage, and insulin action in the tissues (muscle, fat, and liver) of vigorous-intensity exercisers, thereby resulting in a "rebound" in serum VLDL and triglycerides in these groups.
The fact that only high-volume, vigorous-intensity exercise resulted in acute and sustained improvements in HDL cholesterol suggests that the major modifier of HDL metabolism in response to exercise training is likely body habitus variations, such as visceral and subcutaneous fat stores and skeletal muscle mitochondrial capacity, which takes longer to regress posttraining. Conversely, changes in LDL size appear to regress fairly rapidly with cessation of exercise training in all groups. This parallels the relatively rapid reversal of insulin action in all groups. It is likely that these two metabolic parameters are functionally related although the mechanisms whereby these are linked can only be speculative at this point.
Because of the difficulty of conducting exercise detraining studies (individuals who exercise regularly are quite reluctant to stop exercising), relatively few studies have investigated the effects of exercise cessation on maintenance of changes in lipids and lipoproteins following the intentional cessation of exercise training. A literature review revealed 11 studies (7–10, 19, 21, 24, 32, 38, 41, 43) on the effects of exercise detraining or cessation on lipids. Of these 11 studies, 8 studies involved groups of 12 or fewer detraining subjects, and only 2 of the 11 studies were randomized controlled studies, with 1 of these 2 studies reporting on only five detraining subjects. Only three of these studies investigated detraining times that were between 3 and 14 days (9, 10, 21); the others investigated detraining times of 1 mo to 1 yr. These three studies reported conflicting results with regard to lipoprotein changes, perhaps due to the confounding effects as noted above. In a recent extensive review of the topic, Leon and Sanchez (20) concluded that the effects of exercise on lipids are markedly inconsistent. Partially as a result of this inconsistency, in their recommendations for future research, they emphasized the need for large, randomized, controlled studies on this topic and identified a specific need for exercise detraining studies that the present study at least partially fulfills.
There are numerous strengths and limitations to the present study. STRRIDE was a multicenter, randomized controlled study investigating different amounts and intensities of exercise training over a relatively long term of 8–9 mo with multiple detraining time points, allowing for the assessments regarding the acute and sustained effects of the different training regimens. Also, for a study of its type, i.e., with a large number of carefully measured physiological variables in each subject, this is a relatively large study, providing power to measure significant effects between group and outcome. Furthermore, the exercise stimulus was well documented as both the exercise intensity and the number of minutes assigned each week were verified by direct observation and/or recording heart rate monitors.
The results should be viewed also with some cautionary notes. First, this is the first report of its kind, and the findings will be greatly strengthened when they are replicated in other large randomized, controlled studies. Second, the time frame of this study extends only to 2 wk beyond the last training session; generalizations beyond this point, which might be of great interest, will require additional study. Third, the subjects in this study were overweight or mildly obese with mild to moderate dyslipidemia and were middle-aged men and postmenopausal women. Normal weight individuals or individuals with normal lipoproteins would likely respond differently to the exercise intervention, the inactive control period, and the extended detraining period.
In conclusion, the importance of studying the duration of exercise effects on lipoproteins is exemplified by the understanding gained from studying the robust but short-lived effect of exercise on insulin sensitivity, as the knowledge gained has improved exercise prescription (the short-lived benefit suggests daily or near daily exercise is optimal) and our understanding of mechanisms related to this effect. However, the number of time-course studies on the sustainability of exercise-induced lipid benefits is quite limited. The data presented herein show that many of the acute benefits of exercise training on lipoproteins are sustained for up to 2 wk after the cessation of the exercise-training stimulus. Additionally, the findings from this study suggest that there may well be a benefit to individualizing the exercise prescription for wellness around the particular lipoprotein abnormality in order to maximize the desired effects. Specifically, with high-amount training, the significant acute improvements in HDL-C, HDL particle size, and large HDL were sustained at significant levels above baseline for up to 2 wk of detraining. On the other hand, if the lipid abnormality involves hypertriglyceridemia, low-intensity exercise may be the best recommendation, as only the lower, more moderate exercise training intensity resulted in sustained VLDL- and total triglyceride-lowering effects. Finally, as we have reported previously, physical inactivity has numerous, statistically significant, detrimental metabolic consequences. In addition to increased body weight, waist circumference, visceral fat, insulin resistance, and reduced cardiovascular fitness, the observations from the present study are that LDL particle number was significantly increased with significant worsening in small dense LDL, LDL-C, and LDL size over 6 mo of continued sedentary lifestyle in overweight and mildly obese individuals. Notably, all exercise regimens prevented the deterioration in the LDL profile observed in the inactive group, and this beneficial effect was maintained even through 15 days of exercise detraining.
|
GRANTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  Highlights From The Literature Physiology, December 1, 2007; 22(6): 355 - 357. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
