Journal of Applied Physiology
Vol. 82, No. 1, pp. 270-277, January 1997
METABOLISM

Training intensity, blood lipids, and apolipoproteins in men with high cholesterol

Stephen F. Crouse¹, Barbara C. O'Brien², Peter W. Grandjean¹, Robert C. Lowe⁴, J. James Rohack³, John S. Green¹, and Homer Tolson¹

¹ Department of Health and Kinesiology, ² Department of Biochemistry/Biophysics, and ³ Health Sciences Center, College of Medicine, Texas A&M University, College Station, Texas 77843; and ⁴ Health Management Center, Baptist Medical Center, Little Rock, Arkansas 72119 

ABSTRACT

INTRODUCTION

MATERIALS AND METHODS

RESULTS

DISCUSSION

ACKNOWLEDGEMENTS

FOOTNOTES

REFERENCES

ABSTRACT

Crouse, Stephen F., Barbara C. O'Brien, Peter W. Grandjean, Robert C. Lowe, J. James Rohack, John S. Green, and Homer Tolson. Training intensity, blood lipids, and apolipoproteins in men with high cholesterol. J. Appl. Physiol. 82(1): 270-277, 1997.Twenty-six hypercholesterolemic men (mean cholesterol, 258 mg/dl; age, 47 yr; weight, 81.9 kg) completed 24 wk of cycle ergometer training (3 days/wk, 350 kcal/session) at either high (n = 12) or moderate (n = 14) intensity (80 and 50% maximal O₂ uptake, respectively, randomly assigned) to test the influence of training intensity on blood lipid and apolipoprotein (apo) concentrations. All physiological, lipid, and apo measurements were completed at 0, 8, 16, and 24 wk. Lipid data were analyzed via two × four repeated-measures analysis of variance ( = 0.0031). Training produced a significant decrease in body weight and increase in maximal O₂ uptake. No interactions between intensity and weeks of training were noted for any lipid or apo variable, and no between-group differences were significant before or throughout training. Therefore, intensity did not affect the training response. Regardless of intensity, apo AI and apo B fell 9 and 13%, respectively, by week 16 and remained lower through week 24 (P < 0.0003). Total cholesterol fell transiently (5.5%) by week 16 (P < 0.0021) but returned to initial levels by week 24. Triglyceride, low-density-lipoprotein cholesterol, and high-density-lipoprotein (HDL) cholesterol did not change with training. In contrast, HDL₂ cholesterol rose 79% above initial levels by week 8 and 82% above initial levels by week 24 (P < 0.0018); HDL₃ cholesterol fell 8 and 13% over the same training intervals (P < 0.0026). These data show that changes in blood lipid and apo concentrations that accompany training in hypercholesterolemic men are not influenced by exercise intensity when caloric expenditure is held constant.

exercise; hypercholesterolemia; lipoproteins

INTRODUCTION

THERE IS BROAD AGREEMENT that physical activity is associated with a reduced risk of coronary heart disease (CHD) (5, 31) and that at least part of this risk reduction may be due to favorable changes in circulating lipids and apolipoproteins (apos) induced by regular physical exercise (9, 13, 15, 26, 39, 42, 52). On this basis, the inclusion of exercise in the therapeutic regimen for those with hypercholesterolemia and other lipid abnormalities has gained widespread acceptance (16, 32). However, surprisingly little empirical evidence exists to support this practice because healthy hypercholesterolemic individuals seldom have been included as subjects in exercise studies. In a review of the literature, Superko and Haskell (40) identified only two studies in which healthy subjects with moderately elevated cholesterol (224-244 mg/dl) completed an endurance training program. Results were inconclusive; after training either triglyceride (TG) (23) or total cholesterol (TC) (3) concentration was lower, whereas high-density-lipoprotein (HDL) cholesterol (HDL-C) was elevated in the single study in which it was measured (3). More recently, Sutherland et al. (41) reported that TC and low-density-lipoprotein cholesterol (LDL-C) concentrations were higher in hypercholesterolemic men after training, but HDL-C concentration did not change.

Reasons for these diverse findings in hypercholesterolemic subjects are not presently known. It is noteworthy that results of training studies in normocholesterolemic subjects also vary considerably. Generally, it is reported that HDL-C, HDL₂-C, and apo AI are increased while TG concentrations are reduced after training in normocholesterolemic individuals (13, 26, 39, 42, 52), but contrary findings have also been published with surprising regularity (6, 11, 13, 24, 49). Several authors have proposed that study differences in pretraining lipid concentrations, diet, weight loss, training intensity, and weekly volume of training contribute to the disparate findings (13, 43). These factors may be postulated to confound the results of studies in hypercholesterolemic individuals as well.

The issue of training intensity is particularly relevant because moderate-intensity (3-6 metabolic equivalents) training is now generally recommended for promoting health benefits (4, 33), yet it is not clear that this intensity of exercise can induce lipid or apolipoprotein changes consistent with lower CHD risk in either normo- or hypercholesterolemic individuals. It has been reported that a minimum intensity equal to ~75% of maximal heart rate is required to induce favorable HDL-C changes in healthy middle-aged men (39). This finding has not been verified by others (1, 18, 37), and favorable lipid changes have been noted to occur in normocholesterolemic individuals after training at low and moderate intensities (25, 28, 34). Training intensity was either not reported or was poorly quantified in all previous studies involving healthy hypercholesterolemic men (40, 41). Given the absence of published research related to training intensity and lipids in men with above-normal cholesterol, as well as the paucity of information related to general lipid changes that may be expected in these individuals after exercise training, we designed the present investigation to compare the effects of 24 wk of training at moderate and high intensities on blood lipid and apo concentrations in men with elevated cholesterol.

MATERIALS AND METHODS

Subjects

Table 1. Physical characteristics and diet data throughout 6 mo of training in hypercholesterolemic men exercising at high or moderate intensity Baseline 8 wk 16 wk 24 wk Age, yr   Mod 47 ± 10    Hi 48 ± 10  Height, cm   Mod 175 ± 7    Hi 171 ± 5  Weight,a kg   Mod 83.1 ± 15.6  82.8 ± 15.2  81.7 ± 15.3  81.6 ± 15.3    Hi 80.4 ± 9.8  80.1 ± 10.2  79.0 ± 10.0  79.2 ± 10.1  %Fat, %    Mod 27 ± 4  26 ± 5  27 ± 4  27 ± 4    Hi 28 ± 4  27 ± 5  27 ± 3  27 ± 3  Waist-to-hip ratio   Mod 1.02 ± 0.03  1.00 ± 0.03  1.00 ± 0.04  1.01 ± 0.04    Hi 1.02 ± 0.02  1.02 ± 0.02  1.02 ± 0.05  1.02 ± 0.02   O2 max,b l/min   Mod 2.65 ± 0.41c 3.05 ± 0.59d 3.40 ± 0.58e 3.31 ± 0.51e   Hi 2.39 ± 0.43c 2.93 ± 0.54d 3.40 ± 0.50e 3.61 ± 0.43f Total intake, kcal/day   Mod 2,002 ± 441  2,218 ± 617  2,131 ± 559  2,060 ± 730    Hi 1,804 ± 691  1,991 ± 837  1,951 ± 720  1,935 ± 575  CHO, g/day   Mod 210 ± 50  239 ± 67  234 ± 64  224 ± 88    Hi 196 ± 70  218 ± 76  229 ± 76  218 ± 73  Fat, g/day   Mod 84 ± 27  92 ± 31  81 ± 32  89 ± 37    Hi 73 ± 36  86 ± 45  75 ± 41  73 ± 32  Pro, g/day   Mod 82 ± 23  92 ± 26  95 ± 32  77 ± 18    Hi 82 ± 26  83 ± 34  83 ± 29  91 ± 20  Chol, mg/day   Mod 259 ± 157  278 ± 131  318 ± 192  286 ± 120    Hi 261 ± 180  241 ± 215  298 ± 219  282 ± 211 Values are means ± SD for 6 men in each group. Mod, moderate-intensity cycle ergometer training; Hi, high-intensity cycle ergometer training; %Fat, percent body fat; O2 max, maximal oxygen uptake; Total intake, total dietary intake; CHO, dietary carbohydrate intake; Fat, dietary fat intake; Pro, dietary protein intake; Chol, dietary cholesterol intake. a Significant period effect for weight, P  0.05. b Significant group by period interaction for O2 max, P  0.05. c,d,e,f Significant differences between row means, P  0.05.

Procedures

RESULTS

On average, subjects completed 70 of 72 (97%) training sessions over the 6-mo study, and data from subjects who failed to complete at least 90% of the sessions were not included in the analysis. The subjects' physical characteristics and dietary data are presented in Table 1. No significant differences were found between the Mod and Hi groups for height, body weight, waist-to-hip ratios, and percent fat or for any of the dietary variables throughout the 24 wk of training. After 16 wk of training, weight for both groups was significantly lower than baseline values and remained lower through 24 wk (P = 0.0001). A group by period interaction was found for O_{2 max} (P < 0.0013; Table 1, Fig. 1). Both moderate- and high-intensity training elicited significant improvements in O_{2 max} by the 8th wk of the study, but the rate of improvement was greater in the Hi compared with the Mod group (23 and 15%, respectively). This trend for a greater rate of improvement in O_{2 max} with high-intensity training continued throughout the study, and, whereas improvement in O_{2 max} leveled off between weeks 16 and 24 in the Mod group, an additional 6% rise was noted in the Hi group.

The lipid and lipoprotein data for each group are presented in Table 2. No group by period interactions or between-group differences were significant for any of the lipid or apo variables. A significant main effect for training period was found for TC, HDL₂-C, HDL₃-C, apo AI, and apo B concentrations; changes in TG, LDL-C, and HDL-C concentrations after training did not reach statistical significance. Reanalyzing the nonnormal TC data after log₁₀ transformation did not change the results of the statistical analyses. The combined group average (means collapsed across groups) for TC concentration was significantly lower than baseline after 16 wk of training (P < 0.0021) but returned to near-baseline levels by week 24 (Fig. 1). Although total HDL-C did not change with training, HDL₂-C concentration rose (P < 0.0018) and HDL₃-C fell (P < 0.0026) by the 8th wk; these respective changes were maintained through 24 wk of training (Fig. 1). Both apo AI and apo B concentrations were significantly lower than baseline after 16 wk of training and remained lower through the remainder of the study (P < 0.0003; Fig. 1). No lipid or apo changes were correlated with changes in O_{2 max}, body weight, or percent body fat (Table 3).

Table 2. Lipid and apolipoprotein concentrations throughout 6 mo of training in hypercholesterolemic men exercising at high or moderate intensity Baseline 8 wk 16 wk 24 wk TC*   Mod 264 ± 44  256 ± 34  239 ± 31  252 ± 29    Hi 252 ± 17  257 ± 22  245 ± 14  249 ± 20  TG   Mod 207 ± 78  207 ± 124  173 ± 63  183 ± 104    Hi 123 ± 52  127 ± 45  145 ± 61  122 ± 44  LDL-C   Mod 178 ± 44  170 ± 37  163 ± 26  175 ± 34    Hi 180 ± 17  185 ± 23  172 ± 9  180 ± 19  HDL-C   Mod 43 ± 8  46 ± 8  46 ± 10  43 ± 9    Hi 48 ± 9  49 ± 7  48 ± 12  48 ± 11  HDL2-C*   Mod 5.9 ± 3.7  9.4 ± 6.7  9.5 ± 6.0  9.3 ± 5.5    Hi 6.3 ± 4.2  12.6 ± 4.2  11.7 ± 6.0  13.2 ± 9.0  HDL3-C*   Mod 37.8 ± 8.0  36.4 ± 5.9  36.0 ± 6.0  34.0 ± 8.7    Hi 41.5 ± 8.2  36.3 ± 5.4  35.4 ± 7.0  34.4 ± 7.2  Apo AI*   Mod 138 ± 32  140 ± 16  126 ± 13  123 ± 16    Hi 146 ± 19  142 ± 19  132 ± 21  131 ± 24  Apo B*   Mod 109 ± 21  98 ± 16  88 ± 10  90 ± 11    Hi 94 ± 25  97 ± 14  91 ± 13  88 ± 12 Values are means ± SD in mg/dl for 12 men in Hi group and 14 men in Mod group. TC, total cholesterol; TG, triglyceride; LDL-C, low-density-lipoprotein cholesterol; HDL-C, high-density-lipoprotein cholesterol; Apo AI, apolipoprotein AI; Apo B, apoliporprotein B. * Significant period effect for TC, HDL2-C, HDL3-C, apo A-I, and apo B, P < 0.0031 (refer to Fig. 1).

Table 3. Correlation matrix for change scores of variables of interest Variable  TC  TG  LDL-C  HDL-C  HDL2-C  HDL3-C  Apo AI  Apo B  O2 max  0.09 0.21  0.19 0.06 0.16  0.08 0.22 0.21 (0.67) (0.29) (0.36) (0.77) (0.45) (0.71) (0.29) (0.29)  Weight  0.15 0.20  0.07 0.05 0.22  0.21 0.01 0.02 (0.48) (0.34) (0.74) (0.79) (0.29) (0.31) (0.96) (0.93)  %Fat  0.24  0.12  0.14  0.14  0.11  0.02 0.05  0.14 (0.23) (0.57) (0.51) (0.49) (0.58) (0.91) (0.79) (0.49) Values are change scores () computed by subtracting the baseline from 24-wk value and are given as Pearson correlation coefficients with probability in parentheses.

DISCUSSION

Our finding that volume-equivalent training at either moderate or high intensity improves aerobic capacity (O_{2 max}) was not unexpected and is consistent with previous reports in normocholesterolemic men (4, 18, 39). It is of interest that high- compared with moderate-intensity training produced a more rapid rate of improvement and a higher posttraining O_{2 max} after 24 wk (Fig. 1). This is in contrast to the generally held notion that improvements in O_{2 max} are independent of intensity above a minimum threshold, as long as the total energy costs (volume) of training are constant (4). Other notable exceptions besides our own have been published (12, 21, 39). Taken together, the current literature suggests that high-intensity training should be chosen if the rate of improvement in cardiovascular fitness and the highest attainable O_{2 max} are objectives of a training program.

The subjects' average caloric intakes remained stable over the duration of the study, and no significant change occurred in the composition of their diets (Table 1). Thus it is likely that the body weight loss noted (1.4 kg) was in response to the increased caloric expenditure of training. The absence of a significant group (intensity) effect indicated that the weight loss was not affected by intensity. Furthermore, given the caveat that our measurement techniques may have been insensitive to small training-related changes in total or regional body composition, we were not able to detect changes in these anthropometric factors in response to training.

The central finding in the present study was that training intensity had no effect on posttraining changes in lipid and apo concentrations in hypercholesterolemic men. We purposely chose our two training intensities to represent the low- and high-intensity limits commonly recommended to promote cardiovascular fitness (4) and designed our training sessions to ensure an equal volume of training (calories expended) for both intensity groups to isolate training intensity as the primary independent variable of study. The findings show that when the goal of an exercise training program for hypercholesterolemic men is to alter lipid and apo concentrations, no advantage is gained by training at high intensity. This is consistent with current recommendations for promoting public health (4, 33).

We are aware of no other published investigations in men with elevated cholesterol in which exercise intensity was specifically studied. By comparison, Stein and associates (39) reported that 12 wk of cycle ergometer training by healthy normocholesterolemic men (TC  217 mg/dl) induced a significant rise in blood HDL-C and fall in LDL-C only when the intensity was above 75% of maximal achieved heart rate. Unfortunately, these investigators did not control for differences in training volume between intensities, nor did they report the length of time that elapsed between the last training session and blood sampling, either of which could confound the study results. In contrast, Aellen and colleagues (1) reported that HDL-C concentration fell after high-intensity training but rose significantly after training at low intensity. It is difficult to compare their findings with those of others or our own because blood was drawn on nonfasting rather than fasting subjects after exercise was withheld for only 12 h. To further cloud the issue, Gaesser and Rich (18) reported that blood lipid concentrations were unchanged in healthy men after 18 wk of either low-intensity (45% O_{2 max}) or high-intensity (80% O_{2 max}) cycle ergometer training, despite the fact that training at both intensities raised O_{2 max} and reduced body fat.

More recent evidence in older adults (>50 yr) suggests that intensity is less important than the length of time one has trained, if lipid changes are the goal of training. King and associates (27) reported that both high- and low-intensity training (73-88 and 60-70% peak heart rate, respectively) promoted a modest 3 mg/dl rise in HDL-C concentration in previously untrained individuals, but this increase was not apparent until after 2 yr of uninterrupted training. The fact that TC, TG, and LDL-C concentrations were unaltered in this study is consistent with our own findings. Furthermore, significant changes in lipid concentrations have been noted in normocholesterolemic subjects after low- and moderate-intensity training, findings that support the notion that high intensity is not required when lipid changes are the goal of training (25, 28, 34).

The lack of significant changes in traditional lipid risk factors with training in our study was somewhat surprising in light of published research suggesting that pretraining TC, TG, and HDL-C concentrations are correlated with the magnitude of their respective posttraining changes (44) and that physical activity is significantly related to HDL-C levels in hypercholesterolemic men (20). Moreover, although evidence to the contrary exists (6, 13, 24), it is generally believed that exercise training causes favorable changes in blood lipids and lipoprotein cholesterols (13, 26, 27, 42, 44). On this basis, exercise is typically included in the treatment paradigm for those with high-risk lipid profiles (16, 32). However, well-controlled comparative training studies in men with above-average cholesterol are rare and currently inconclusive. TC, LDL-C, TG, and HDL-C concentrations are variously reported to be raised, lowered, or unchanged in hypercholesterolemic men after exercise training (3, 23, 25, 40, 41). Conclusions that can be drawn from these published studies are limited by the absence of a control group (3), differences among studies in mode and volume of training, subject pretraining lipid levels, diet control, and genetic basis for the elevated cholesterol, as well as by a failure to control for the acute effects of the most recent training session (23, 25, 41).

Our null findings should be interpreted in light of previous research suggesting that beneficial changes in the lipid profile with training may not be independent of weight loss (48, 51) or the duration of the training program (27). It is possible that training beyond 6 mo combined with more substantial weight loss may have produced different results in the hypercholesterolemic subjects in our study. Furthermore, given that a threshold of energy expenditure may exist for producing lipid changes in normocholesterolemic men (equivalent to running 8 mi/wk) (51), it is conceivable that the threshold may be higher in hypercholesterolemic men, at least above the 1,050 kcal/wk energy expenditure we used in our study. Recent evidence suggests that exercise should be considered just one component of a multifaceted risk-reduction strategy. Regular exercise combined with a low-fat low-cholesterol diet has proven to be a useful combination to raise HDL-C in moderately overweight normocholesterolemic men (53). At present the effectiveness of a similar combined strategy in hypercholesterolemic men has not been tested.

Although any or all of these factors may impact lipid and exercise studies, we propose a simpler explanation for our null findings, particularly with respect to the absence of a significant TG reduction, based on research related to the timing of blood sampling procedures after exercise. Holloszy et al. (23) reported that an apparent reduction in TG concentration with training in hypercholesterolemic men was, on closer inspection, a transient response to the previous training session that persisted for at least 44 h. This early finding of a transient (acute) response to exercise has been corroborated by others in normocholesterolemic men (13) and by our laboratory in untrained hypercholesterolemic men (10). In the study from our laboratory, the nadir of TG concentration occurred at 24 h, after which TG rose ~5% toward preexercise levels by 48 h, suggesting that the effect of exercise was dwindling (10). Indeed, a review of the related literature suggests that the transient lipid response, including that for TG and HDL-C, generally lasts 24-48 h after moderate amounts of exercise in normocholesterolemic individuals and perhaps up to 72 h after maximal exertion, such as a marathon (13). Blood specimens for analysis were collected ~60-72 h after exercise in our present study. On the basis of previous work by ourselves and others (10, 13), we believe it is reasonable to infer that by this time concentrations of traditionally measured lipid risk markers had returned to near preexercise values and, therefore, showed no apparent response to training. Unfortunately, with the data we report in this present communication we cannot confirm this hypothesis, nor can we determine whether the transient lipid response is altered by training. We can only conclude that compared with pretraining values, lipid concentrations measured at least 60 h after the last exercise session were not significantly lower after 2, 4, or 6 mo of endurance training. The current literature offers little help in resolving this issue. Blood was collected for analysis either without regard to the most recent training session(s) or within 48 h of exercise in most published studies linking exercise training to beneficial changes in lipids and apos, a flaw in research design we believe has led, in large part, to confusion with regard to the transient vs. chronic effects of exercise. We are aware of no published studies specifically designed to test the influence of training on the transient lipid response to a single session of exercise. This lack of research leaves open the possibility as suggested by Haskell (22) that a substantial portion of the health benefit often attributed to chronic exercise training may be, in fact, a manifestation of the lipid response to the most recent episode of exercise or an accumulation of benefit from several repeated exercise sessions. If such is the case, the frequency of exercise would be a critical component of the exercise prescription process when beneficial lipid changes are the goal of training.

Although traditionally measured lipids and lipoproteins were essentially unchanged, HDL₂-C rose 82% and HDL₃-C fell 13% by the 24th wk of training, demonstrating subtle changes in HDL metabolism not evident when HDL-C alone was measured. The major portion of these changes occurred by the 8th wk (Fig. 1), suggesting that HDL metabolism changes rapidly in response to training. To our knowledge, no comparative data from hypercholesterolemic men have been published. Findings from studies in normocholesterolemic men are presently inconclusive with training variously accompanied by an increase, decrease, or no change in the blood concentration of one or both of these HDL-C subfractions (29, 34, 35, 38, 42, 48, 51, 52). To explain these discrepancies in study results, it has been proposed that exercise-induced changes in the major HDL subfractions are dependent on weight loss (47). Our results do not agree; neither the change in weight nor in body fat was associated with changes in HDL₂-C, HDL₃-C, or any other lipid-related variable measured in our study (Table 3).

We noted an 11% fall in apo AI concentration, an outcome contradictory to reports that either no change (42, 51) or an increase (26, 38, 46) in apo AI concentration accompanied training in normocholesterolemic men. We have no adequate explanation for this finding. Normally, increases with training are thought to be due to augmented apo AI synthesis combined with reduced catabolism (42). Although we lack direct apo AI kinetic data, our results suggest that either synthesis was impaired or catabolism increased in our hypercholesterolemic subjects over the 24-wk training period. The fact that apo AI synthetic rates are reportedly lower in hypertriglyceridemic compared with normal subjects (14) suggests that apo AI synthesis may be related to lipid status. At present, corresponding data are lacking for hypercholesterolemia.

The reduction in apo B that accompanied training in our study has not been previously reported for hypercholesterolemic men, and few comparative training studies in men with normal cholesterol have been published. Wood et al. (51, 53) reported that apo B was reduced after 1 yr of weight loss by dieting combined with exercise but not after 1 yr of exercise alone. Dietary caloric intake did not change in our study; neither weight loss nor changes in body fat were substantial, and these factors were not correlated with reductions in apo B (Table 3). Thus our data suggest that significant reductions in this putatively atherogenic apo can occur in hypercholesterolemic men in the absence of substantial changes in body weight and composition.

We conclude that in hypercholesterolemic men changes in lipid and apo profiles that accompany traditional endurance training are not dependent on intensity, at least not for calorically equivalent training at intensities equal to 50 and 80% of O_{2 max}. However, the absence of a control group limits the conclusions that can be drawn from the training aspect of our study, and we cannot rule out the potential confounding influence of seasonal variation. Given this caveat, our results show that 24 wk of exercise training at a caloric cost of 1,050 kcal/wk is accompanied by significant changes in HDL₂-C, HDL₃-C, apo AI, and apo B in hypercholesterolemic men. Whether such changes alter atherosclerotic risk in this population remains to be shown.

ACKNOWLEDGEMENTS

We are grateful to Jana Crouse and Susan Lee for their assistance in the preparation of this manuscript and to Drs. Nicolaas Pronk and Dennis Jacobsen for their technical assistance in data collection.

FOOTNOTES

Address for reprint requests: S. F. Crouse, Applied Exercise Science Laboratory, Dept. of Health and Kinesiology, Texas A&M Univ., College Station, TX 77843.

Received 21 March 1996; accepted in final form 6 September 1996.

REFERENCES