INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

CARBOHYDRATE (CHO) supplementation during moderately intense exercise [~70% maximal oxygen consumption (O_{2 max})] has been shown to consistently improve exercise time to fatigue by 24-32% in men (6, 8, 40). Fewer data concerning the beneficial effects of CHO supplementation on endurance performance in women are available. Tarnopolsky et al. (37) reported that the ingestion of a high-CHO diet did not influence endurance performance in the women studied. These authors and others (14, 37, 38) suggest that women oxidize lipid to a greater extent and CHO to a lesser extent than do men during moderately intense exercise. Consequently, it is possible that CHO supplementation during exercise may not have the same effects in women as in men.

Differences between men and women in substrate utilization during moderately intense exercise may be related to the reproductive hormones estrogen and progesterone. Elevated concentrations of estrogen and progesterone that occur during the luteal (Lut) phase of the menstrual cycle have been associated with enhanced glycogen storage (14, 32) and fat utilization during rest and exercise (14). Although this shift in substrate metabolism may be beneficial during prolonged low-intensity exercise (12), higher intensity endurance exercise in the absence of glucose feedings (7) may be adversely affected.

Although several investigations have described variations in athletic performance during different phases of the menstrual cycle (5, 22, 23), there is no consistent evidence to suggest that the potential shift in substrate utilization described above has any effect on endurance performance during prolonged exercise. Furthermore, the effects of menstrual cycle phase on the potential positive effect of CHO supplementation on endurance performance during prolonged exercise are unknown. Consequently, the purpose of this investigation was to determine whether menstrual cycle phase influences the effect of CHO supplementation on substrate metabolism and fatigue during prolonged exercise.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects. Eleven healthy, moderately trained women volunteered to participate in the study. All procedures were approved by the University's Institutional Review Board for the Protection of Human Subjects in Research, and participants provided their informed consent. Subjects completed a medical history questionnaire to rule out any contraindications to participation and an activity questionnaire to document current exercise patterns. Although subjects were not trained cyclists, all had previous experience with endurance cycling. The women were eumenorrheic and were not using oral contraceptives (OCs). Cycle phases were validated by measurement of serum estrogen and progesterone. Two women were excluded from the study due to Lut-phase progesterone values below 16 nmol/l (20).

Experimental protocol. Experimental sessions were conducted on four separate occasions, twice during the follicular (Fol) phase (cycle days 1-8) and twice during the Lut phase (cycle days 19-24) of the menstrual cycle. Before each session, subjects were asked to refrain from caffeine and alcohol ingestion for at least 12 h and to refrain from strenuous exercise for 24 h.

1

1

Physiological measures. Heart rates (beats/min) were measured continuously with a heart rate monitor (Vantage XL, Polar) and were recorded every 10 min. Blood pressure was monitored by auscultation using a sphygmomanometer every 20 min during the trials. Mean arterial pressure (mmHg) was calculated from systolic and diastolic pressures (pulse pressure + diastolic pressure). O₂ was measured for a 5-min period every 20 min during the trials by using the system described in Subjects. Total body water loss was calculated from the change in body weight corrected for ingested fluid and urine volume.

Psychological measures. To assess the potential influence of CHO supplementation on subjective feelings of fatigue, ratings of perceived exertion (RPE) (4) were recorded every 30 min.

Blood analyses. Aliquots of plasma were drawn from centrifuged blood containing either heparin (amino acids and cortisol) or EDTA [free fatty acids (FFA) and glucose]. Blood samples for determination of plasma catecholamines [norepinephrine (NE) and epinephrine (Epi)] were collected into chilled heparinized vacutainers containing 100 µl of a reduced-glutathione and EGTA solution before centrifugation. Aliquots of serum were drawn for estradiol and progesterone. Samples were frozen at 70°C for subsequent analyses.

Statistical analyses. Data were evaluated for differences among menstrual cycle phases, drink treatments, and time by using a three-way (phase × drink × time) repeated-measures analysis of variance (SuperANOVA, Abacus Concepts, Berkeley, CA). Significance was set a priori at the P < 0.05 level. When significant main effects were observed, standard contrast procedures were performed by using least squares means. For all values, means ± SE are reported.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Physical characteristics and reproductive hormone data for the nine subjects included in the study are presented in Table 1. Although these women were not trained athletes, their percent body fat and percent O_{2 peak} values placed them in the "excellent" and "superior" categories, respectively, for their age group (1).

Cycling time to exhaustion at 70% O_{2 peak} for all four trials are shown in Fig. 1. A significant increase in endurance time was observed during CHO compared with Pl trials (P < 0.05); menstrual cycle phase had no effect. Improvements in endurance time during CHO treatments were similar for Fol (14.4 ± 8.5%) and Lut (11.4 ± 7.1%) phases. Only one woman failed to improve her endurance time with CHO treatment in both cycle phases.

View larger version (40K): [in this window] [in a new window]   Fig. 1.   Exercise time to fatigue during cycle ergometry at 70% peak O2 consumption in 9 trained women. Values (means ± SE) are for placebo (Pl) and carbohydrate (CHO) trials over both phases of the menstrual cycle. * Significant difference between drink treatments (Pl vs. CHO), P < 0.05.

Plasma FFA and glucose are shown for all four trials over the course of exercise in Fig. 2. Main effects were noted in FFA responses for drink treatment and time (P < 0.05). In addition, the changes in FFA over time were influenced by drink treatment (P < 0.05) and menstrual cycle phase (P < 0.05). As shown in Fig. 2A, FFA increased above baseline at 90 min of exercise and continued to increase until fatigue (90 min < 120 min < fatigue; P < 0.05) for both drinks. This rise in FFA was significantly greater during the Pl compared with CHO trials (P < 0.05). At 120 min of exercise and fatigue, FFA levels were significantly lower during Lut compared with Fol phases (P < 0.05).

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Plasma free fatty acids (FFA; A) and glucose (B) over the 4 exercise trials: CHO, Pl, luteal phase (Lut), and follicular phase (Fol). * Significant difference between drink treatments (Pl vs. CHO), P < 0.05.  Significant differences over time, P < 0.05. § Significant interaction between drink and time, P < 0.05.  Significant interaction between menstrual phase and drink, P < 0.05.

As expected, plasma glucose was significantly greater during CHO compared with Pl trials (P < 0.05). No differences were observed between menstrual cycle phases (Fig. 2B). Although the change in glucose over time was not significant (P > 0.05), a drink by time interaction was evident (P < 0.05). From 60 min of exercise until fatigue, plasma glucose during CHO was significantly greater than during Pl trials (P < 0.05). At fatigue, plasma glucose was elevated above baseline values during CHO trials (4.78 ± 0.21 vs. 4.48 ± 0.15 mmol/l; P < 0.05). In contrast, a significant reduction from 4.52 ± 0.12 to 3.74 ± 0.26 mmol/l (P < 0.05) occurred during Pl trials.

The effect of menstrual cycle phase and CHO supplementation on plasma amino acids during exercise are shown in Table 2. Alanine, glutamine, proline, and isoleucine were all reduced during Lut compared with Fol phases (P < 0.05). CHO ingestion resulted in significantly lower plasma levels of tyrosine, valine, leucine, isoleucine, and phenylanine and in significantly elevated levels of plasma alanine compared with Pl trials (P < 0.05). Exercise at 70% O_{2 peak} resulted in decreases over time in plasma alanine, glutamine, valine, and proline (P < 0.05). A significant interaction (P < 0.05) was noted between cycle phase and drink treatment for plasma isoleucine and valine. Values were higher during Fol compared with Lut phases in the Pl trials, whereas minimal phase differences were noted for the CHO trials.

CHO resulted in significantly lower values for tryptophan (Table 2) by 120 min of exercise (P < 0.05). There was a trend for a drink by time interaction for tryptophan, although this was not significant (P = 0.10).

Plasma cortisol, insulin, NE, and Epi responses are shown in Fig. 3 for both drink treatments over time. Because neither menstrual cycle phase main effects nor phase interactions with drink or time were observed, data for Fol and Lut phases were combined for each drink treatment. Plasma cortisol increased over time for both drink treatments (P < 0.05; Fig. 3A). Although an overall drink effect was not significant, an interaction (P < 0.05) was found between drink treatment and time; cortisol was higher at 60 min of exercise during CHO compared with Pl trials (357 ± 26 vs. 274 ± 28 nmol/l). At fatigue, cortisol was suppressed during CHO (486 ± 57 nmol/l) compared with Pl (625 ± 59 nmol/l). As shown in Fig. 3C, exercise resulted in a significant reduction (P < 0.05) in plasma insulin. Plasma insulin levels were greater during CHO compared with Pl trials (116.6 ± 16.5 vs. 93.9 ± 8.3 pmol/l) by 60 min of exercise. Plasma NE increased over time until 120 min of exercise for both drink treatments (0 min < 60 min < 120 min; P < 0.05; Fig. 3B). Values at fatigue (1,516 ± 162 pmol/l) were not different from those at 120 min of exercise (1,362 ± 152 pmol/l). Exercise at 70% O_{2 peak} resulted in an increase in Epi for both drink treatments (P < 0.05; Fig. 3C), although CHO supplementation significantly suppressed the Epi response from 120 min (P < 0.05). Plasma Epi at fatigue was over twofold greater during Pl compared with CHO trials (589 ± 77 vs. 253 ± 36 pmol/l).

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Plasma hormone responses over time for Pl vs. CHO trials. A: cortisol. B: norepinephrine. C: insulin. D: epinephrine. * Significant difference between drink treatments (Pl vs. CHO), P < 0.05.  Significant differences over time, P < 0.05. § Significant interaction between drink and time, P < 0.05

Cardiorespiratory and hemodynamic responses averaged over each 30-min period during the trials are shown in Table 3. Neither menstrual cycle phase nor drink treatment altered these responses during 70% O_{2 peak} exercise. Responses for respiratory exchange ratio during the course of exercise are displayed in Fig. 4. Neither drink nor menstrual cycle phase had an impact on RPE.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Respiratory exchange ratio (RER) responses during exercise for Pl vs. CHO trials. @ Significant differences over time, P < 0.05.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The purpose of this investigation was to determine whether menstrual cycle phase influences the effect of CHO supplementation on performance and related plasma substrates during prolonged exercise. The results of this investigation indicate that exercise time to fatigue during prolonged exercise at 70% O_{2 peak} is equally improved by CHO supplementation (0.6 g CHO · kg body wt¹ · h¹) during the Fol and Lut phases of the menstrual cycle.

Although the improvements in endurance performance with CHO supplementation (14% during the Fol phase and 11% during the Lut phase) are consistent with the trends observed by other investigators, the magnitude of improvement is less than that observed in men under similar conditions. For example, Wright and colleagues (40) observed a 32% increase in exercise time to fatigue during cycling at 70% O_{2 max} when CHO was supplied at a rate of 0.6 g CHO · kg body wt¹ · h¹. Similarly, Davis and colleagues (8) observed a 24% increase in exercise time to fatigue during cycling at ~68% O_{2 max} when CHO was supplied at a rate of 0.6 g CHO · kg body wt¹ · h¹. Although these potential differences between men and women are intriguing, subtle differences in exercise intensity or training status cannot be discounted as potential reasons for the lower relative improvement in performance observed here in women. An appropriate comparison of the effects of CHO supplementation on endurance performance between men and women can only be made by an investigation that includes both genders.

Numerous investigations have addressed the differences in physical performance during various phases of the menstrual cycle [see Lebrun (22) for review]. The results of these investigations have greatly varied. For example, Jurkowski and co-workers (18) observed nearly a 100% greater exercise time to fatigue during the Lut phase (2.97 ± 0.63 min) compared with the Fol phase (1.57 ± 0.32 min) during cycling at 85-90% O_{2 max}. Similarly, Nicklas and colleagues (32) observed a small but significantly greater exercise time to fatigue during the Lut phase (139.2 ± 14.9 min) compared with the Fol phase (126 ± 17.5 min) during cycling at 70% O_{2 max}. Although the two investigations cited above suggest that endurance performance is greater in the Lut phase, the bulk of the investigations in this area report no difference in endurance performance between the menstrual cycle phases (5, 22). In the present investigation, the finding that exercise time to fatigue was similar in the Lut and Fol phases during Pl and CHO conditions is in agreement with these latter studies.

Differences in endurance performance during the phases of the menstrual cycle have been postulated to be due to differences in substrate availability and metabolism. Elevated concentrations of estrogen and progesterone during the Lut phase have been associated with altered muscle glycogen and FFA utilization during rest and submaximal exercise (37, 38). These differences in substrate utilization may influence performance during prolonged exercise by altering the availability of muscle glycogen. For example, Lavoie and colleagues (21) found plasma glucose was lower and plasma FFA was higher during cycling at 63% O_{2 max} for 90 min during the Lut phase. In the present investigation, menstrual cycle phase appeared to have no effect on plasma glucose and FFA during exercise. Although our results differ from those described by Lavoie and co-workers (21), they are consistent with several other reports (5, 19, 32). For example, Nicklas and colleagues (32) found blood glucose, plasma FFAs, and respiratory exchange ratio to be similar during the Lut and Fol phases of the menstrual cycle during cycling at 70% O_{2 max}. The lack of consistency of results among investigations may be related to small differences in exercise intensity. In an investigation completed by Hackney and co-workers (14), lipid utilization and oxidation were found to be greater during the Lut phase at exercise intensities equal to 35 and 60% O_{2 max}; however, differences between the Lut and Fol phases were not present at 75% O_{2 max}. Support for the evidence that menstrual phase differences in substrate utilization during exercise varies according to exercise intensity is seen in recent work by Friedlander et al. (11). Results from these investigations indicate that, whereas glucose use during exercise is directly related to exercise intensity in women, FFA use is not. Because the exercise intensity used in this investigation (70% O_{2 max}) was at a level at which CHO metabolism may become exponentially more important to energy production than FFA oxidation, it would be inappropriate to extend our findings to lower exercise intensities. It is important to note, however, that it is unlikely that our subjects would have been able to exercise for >2.5 h during Pl if FFA oxidation was not playing an important role in energy production.

The perception that substrate availability and utilization differs during the Lut and Fol phases of the menstrual cycle has been perpetuated by findings from investigations that describe the effects of OC use on metabolism and plasma substrates (2, 35). Because OC use elevates plasma estrogen levels, and estrogen levels are higher during the Lut phase of a "normal" menstrual cycle, OC use and the Lut phase are often expected to have similar metabolic results. Although these investigations are extremely valuable in the attempt to understand the metabolic events associated with the specific population under examination, these findings should not be extrapolated to normally cycling women. For example, an investigation completed by Ruby and associates (35) suggests that application of a transdermal estrogen patch for 72 and 144 h in amenorrheic women does not affect "whole body" lipid metabolism during 90 min of treadmill running at 65% O_{2 max}. The results of this investigation are extremely important when women with amenorrhea are considered; however, extrapolation to women who are normally cycling is inappropriate for two reasons. First, the pharmaceutical treatment increased estradiol to a level that was >10-fold lower than that see in normally menstruating women. Second, the physiological events that result in women becoming amenorrheic and/or the ensuing state of hypoestrogenism could very well result in adaptations that would impact the metabolic consequences of increasing and decreasing estrogen levels in the blood.

Differences in plasma levels of glucose and FFA were observed in this investigation because of the drink treatment. Specifically, blood glucose levels were maintained and increases in FFA were blunted during the CHO trials. These differences are consistent with those described in men exercising under similar experimental conditions (8, 31, 40). One previous investigation has described the effects of CHO supplementation on plasma glucose and FFA levels during various phases of the menstrual cycle. In this investigation, Bonen and colleagues (3) provided women with 1.5 g/kg of CHO 20 min before 1 h of treadmill walking (30 min at 40% O_{2 max} and 30 min at 80% O_{2 max}). Interestingly, whereas plasma glucose levels were not found to be affected by CHO supplementation or menstrual cycle phase, plasma FFA levels were observed to be lower when CHO was provided before exercise in the Lut phase. The results of this investigation by Bonen and colleagues appear to be contradictory to those investigations (14, 21) that have observed greater plasma FFA levels while exercising during the Lut phase. An explanation for these various results does do not appear to be readily apparent; however, it is possible that the procedures utilized by Bonen and colleagues (CHO supplementation before exercise) precipitated a unique physiological response that warrants further study.

Menstrual phase differences were observed for several amino acids. Plasma levels of alanine, glutamine, proline, and isoleucine were all lower in the Lut phase compared with the Fol phase. These general results are consistent with those of Moller and associates (30), who found the sum of neutral amino acids to be 10% lower in the Lut phase compared with the Fol phase. Interestingly, they only found tyrosine to be significantly lower in the Lut phase. Although glutamine and proline were not measured in the investigation investigation by Moller et al., alanine and isoleucine appeared to follow this trend. This difference in plasma levels of neutral amino acids may be related to estrogen and progesterone levels, which have been found to have a catabolic effect on amino acids (30). We also found an effect of menstrual cycle phase on plasma levels of isoleucine, leucine, and valine. Specifically, plasma levels of these amino acids were higher during the Fol-Pl trial than the Lut-Pl trial. It is unclear why this difference was not also observed in the CHO trials. The subsequent effect of these differences in plasma amino acids as a result of menstrual cycle phase on metabolism and related bodily functions is unknown. Plasma amino acids play important roles in a vast spectrum of functions, including production of various neurotransmitters (7) and CHO metabolism (39).

Prolonged exercise in this investigation resulted in reductions in plasma alanine, glutamine, valine, and proline. Similar findings have been described by other authors in men (13, 15, 34, 39). For example, Graham and colleagues (13) observed decreases in plasma levels of alanine and proline after 3 h of one-legged knee extensor exercise at 60% O_{2 max} in trained men. Van Hall and associates (39) found significant reductions in plasma valine levels after 90 min of one-legged knee extensor exercise at 60-65% O_{2 max} in trained men with normal muscle glycogen levels. Furthermore, Rennie and co-workers (34) describe significant decreases in plasma glutamine and alanine as a result of 3.5 h of cycling at 50% O_{2 max}. Declines in plasma amino acids during exercise may be the result of deamination of existing amino acids or reduced synthesis of amino acids. It is generally believed that these "extra" carbon skeletons are then used for energy production (15). Interestingly, it appears that the present investigation is the first to observe simultaneous declines in the four amino acids: alanine, glutamine, valine, and proline. Mechanisms for this finding and its importance are unclear. Whether these findings are the result of the slightly greater exercise intensity used in the present study and/or the use of two-legged cycle ergometry vs. one-legged exercise requires further study. It also appears that this is the first investigation to observe this response in women.

In this investigation, CHO supplementation resulted in significantly lower plasma levels of tryptophan and the branched-chain amino acids (BCAA; leucine, isoleucine, and valine) by 120 min of exercise. These results are generally consistent with those described by Davis and colleagues (8). The attenuation in the rise in plasma tryptophan levels is believed to be the result of attenuation in plasma FFA levels as a result of the CHO supplementation. Increasing levels of plasma FFA result in increased plasma levels of tryptophan by displacing tryptophan from albumin (7, 8). Conversely, declines in plasma BCAA levels during CHO supplementation and exercise are thought to be due to the maintenance of plasma insulin levels during exercise (7, 29).

Along with the BCAA, CHO supplementation also resulted in lower plasma levels of tyrosine and phenyalanine. Since movement of these amino acids into muscle and liver can also be enhanced by insulin (8, 27), reduced levels of these two amino acids during the CHO trials are believed to be a result of the same mechanism related to the BCAA.

CHO supplementation also influenced the hormonal responses to exercise. Specifically, CHO supplementation during prolonged exercise attenuated increases in plasma cortisol and Epi and decreases in plasma insulin. These results are generally consistent with previous investigations (26, 28, 31); however, the magnitude of decline in plasma insulin levels during CHO supplementation appears to be greater in women compared with the decline previously described in men. For example, Murray and colleagues (31) observed when subjects received a Pl drink that plasma insulin levels decreased up to ~30% during 120 min of cycling at 65-75% O_{2 max}. In comparison, when subjects were provided a CHO beverage (26 g/h), plasma insulin levels decreased ~15% during the first 20 min of exercise and tended to approximate baseline levels during the remaining 100 min. In this investigation, we observed that plasma insulin levels decreased ~50% during the first 60 min of CHO and Pl trials. The significance of this finding is unclear. Several investigations have been completed that describe differences in insulin sensitivity between men and women (9, 10). In one investigation, Donahue and associates (9) observed that premenopausal women have a greater insulin response to an oral glucose tolerance test than do men. In another investigation, these same authors found that glucose uptake at a constant insulin level was greater in women than in men after adjustments were made for differences in body composition (10).

In conclusion, the results of this investigation indicate that the performance-enhancing effects of CHO supplementation during prolonged exercise at 70% O_{2 max} are not influenced by menstrual cycle phase. Furthermore, menstrual cycle phase did not alter endurance performance or any hormonal and metabolic factors that could influence endurance performance.