Introduction
Individuals who expend significant amounts of energy, such as athletes, must be aware of their energy requirements and should be diligent about matching these levels with adequate dietary intake to avoid a state of negative energy balance. Prolonged negative energy balance increases the risk of overtraining syndrome, stress fractures, and relative energy deficiency in sport (RED-S), as well as compromises the immune system, athletic performance, training adaptations, and recovery (10,18,27). Alarmingly, the prevalence of negative energy balance among athletes is high, whereby studies suggest that adolescent- and collegiate-aged populations may be particularly vulnerable (5,27,29). Another measure of energy status is energy availability. Energy availability measures the amount of energy available to devote to basic metabolic functions after the energy expended from physical activity and exercise has been taken into account (11). Energy availability is calculated by subtracting exercise expenditure from energy intake and then dividing this difference by kilograms of fat-free mass (FFM) (25). Low energy availability stems largely from inadequate energy intake and can contribute to the development of RED-S (17,25).
For a diverse group of athletes whose sport is intermittent in nature and requires travel to and from games across a period of several months, maintaining appropriate energy balance throughout a competitive season can prove to be a challenge (25). Energy intake requirements can change on a weekly and sometimes daily basis because of varying factors such as the number of games and the time and intensity devoted to practice (29). Failure to adjust energy intake properly may lead to changes in both fat mass and lean body mass while also compromising an individual's ability to adequately recover, which may affect an athlete's health and performance (9,30). Although athletes typically meet protein and fat intake recommendations throughout a season, their calorie and carbohydrate intake levels are often below recommended levels (9,20). The current body of literature reports equivocal findings regarding changes in energy availability, body composition, and other performance measures throughout entire seasons (3,12,25,29,31). More specifically, low energy availability has been found to be especially prevalent during preseason and midseason, which suggests that athletes fail to consume enough nutrients in their diets at these times (25,30). Such behaviors could be particularly problematic for female athletes because low energy availability has been shown to increase their risk of RED-S, which is associated with a myriad of negative health effects including menstrual cycle dysfunction and an elevated risk of low bone mineral density (21). By contrast, significant increases in female basketball players' FFM over a season have also been reported, which one could interpret as an adequate delivery of nutrients (25,30). Although other studies have addressed energy balance over an entire athletic season, most reports only complete measurements taken before and after the season. This measurement frequency may be insufficient to accurately evaluate changes throughout a season (3,9,20,31).
Finally, another challenge relates to the broad and sometimes nebulous approach used to assigning dietary intake levels for athletes. The need to research the change in energy requirements throughout an athletic season in more detail has been expressed by many publications, largely because sport nutrition recommendations vary based on individual metabolic rates, training loads, competition schedule, and position-specific demands (9,29,31). Therefore, the purpose of this project was to evaluate energy balance and determine the extent to which daily energy needs change across an entire women's collegiate basketball season.
Methods
Experimental Approach to the Problem
This was an observational study that spanned over an entire women's collegiate basketball season. Data related to activity energy expenditure, total daily energy expenditure (TDEE), energy intake, aerobic fitness, resting energy expenditure (REE), body composition, and questionnaires to assess sleep habits and recovery were collected during 5 phases throughout the season with each phase separated by approximately 1 month. Subjects were required to wear monitors that collected combined heart rate and accelerometry data over 4 consecutive days (2 weekdays and 2 weekend days), during each of the 5 phases throughout the season (total of 20 days). Dietary logs were also completed over the same 4-day period that activity monitors were worn. Phase I consisted of heavy practicing for 2–3 hours, separated by 1 or 2 game days of nonconference basketball games. On the Sundays of most weeks during this time window, subjects did not participate in any athletic related activities. This was defined as their off day. Phase II was similar in structure, except all games but 1 were in conference league play. Phase III was characterized by a decline in practice time from 2 to 3 hours to an hour and 30 minutes at most. Also in phase III, postseason conference tournament play took place, extending the season by an additional 3 games and 5 practices. Phases IV and V were characterized by off-season workouts. No games were played during this time. Weekdays contained 1 basketball specific training day paired with a conditioning workout of either aerobic or anaerobic training. The following day included resistance training for an hour. On the weekends, subjects were relieved of all athletic-related duties.
Subjects
Members of a National Collegiate Athletic Association (NCAA) Division II women's basketball participated in this study. Before initial data collection periods, study details and participation requirements were explained. For inclusion, subjects had to be current members of Division II women's basketball programs and regularly participating in all team activities. A total of 13 female athletes (mean ± SD: 19.8 ± 1.3 years [range 18-23 years], 173.9 ± 13.6 cm, 74.6 ± 9.1 kg, and 27.1 ± 3.2% fat) completed the study protocol. Written informed consent was provided, and all procedures were approved by IntegReview (Protocol # 964572, approval date: October 16, 2016).
Procedures
Body Composition
At the beginning of the data collection period, body composition assessments were completed using dual-energy x-ray absorptiometry (DEXA) for basic demographic information and for calculation of energy availability. Body composition and body mass measurements were only completed at 1 time point because of concern from the sport coaches and other support staff about preoccupation with body mass or body image among the players. To standardize pretesting conditions before completing the DEXA, subjects were instructed to arrive after observing an 8-hour fast of energy-containing food and fluid and avoiding exercise for at least 24 hours (22). Subjects were instructed to ingest water ab libitum to avoid dehydration; however, a hydration assessment was not performed. Calibration procedures were completed each day before testing, and all DEXA scans were completed using a Hologic QDR Discovery A (HOLOGIC, Bedford, MA) and analyzed using its accompanying software (Hologic APEX Software, Version 4.5.3; HOLOGIC) to determine whole-body levels of bone, fat, and FFMs along with body fat percentages. All results were calculated by the software's proprietary algorithm based on the data from the 2008 NHANES survey. All DEXA scans were performed and analyzed according to the manufacturer guidelines by the same 2 laboratory investigators. Test-retest reliability of performing this test on 20 healthy, college-aged individuals on our equipment yielded intraclass correlation coefficients of ≥0.998.
Maximal Oxygen Consumption
For descriptive purposes, a maximal oxygen consumption test was performed at approximately the midpoint of the investigation. During the test, expired gases were continuously collected by a head set containing a 2-way valve (Hans Rudolph) and a nose clip worn by subjects and analyzed by a metabolic cart using indirect calorimetry (TrueOne 2400; ParvoMedics, Sandy, UT). On each day of testing, the metabolic cart ran for 30 minutes before calibrations were completed against measured gas concentrations and flow rate. A calibration was not accepted, unless the new calibration point was within 2% of the previous calibration point. Subjects were first outfitted with a Polar FT1 heart rate monitor (Polar Electro, Inc. Bethpage, NY) with accompanying chest strap with the electrodes first being moistened with tap water and worn at the level of the xiphoid process. A graded treadmill test was completed using a Woodway Desmo Treadmill (Waukesha, WI) interfaced to the metabolic cart with all grade and speed progressions being automatically controlled by the metabolic cart. According to the procedures of Camic et al. (8), the test began at a speed of 3 mph and 0% grade. Every 2 minutes, the speed was subsequently increased by 1 mph, until a speed of 9 mph was reached. At this point, the grade was increased by 2.5%, until the person reached volitional fatigue. The test was considered maximal if any 2 of the 3 events occurred: the subject reached a respiratory exchange ratio ≥1.10, recorded heart rates were within 10 b·min–1 of the individuals age-predicted maximal heart rate using the formula of Tanaka (Maximum heart rate = 210 − [0.7 × age] (33)), and V̇o2 changed less than 250 ml·min−1 for 2 consecutive sampling points.
Resting Energy Expenditure
All REE measures were completed using a ParvoMedics TrueOne 2400 metabolic measurement system. Each morning, the indirect calorimetry system was calibrated by a research team member to ensure that variations of measured oxygen and carbon dioxide as well as flow rate were less than 2% different than the previous calibration. All subsequent tests were completed in an isolated, thermoneutral room with the lights illuminated. A blanket was provided, and a clear plastic hood and drape were placed over each subject's head and shoulders. The flow rate on the dilution pump was set to maintain approximately 0.8–1.2% carbon dioxide in the expired gases. Once an appropriate flow rate was established, study subjects remained awake and motionless in a supine position for 20–25 minutes. Data were collected minute-to-minute. The recorded data were visually inspected, and a 5-minute window where V̇o2 (in L·min−1) changed less than 5% was identified before being averaged and used for analysis. From this group of data, REE values (in kcal·d−1) were calculated, and the average of all data points was computed. All tests were completed under fasted conditions, whereby subjects were instructed to fast from all energy-containing foods and fluids for a minimum of 8 hours before the test. All measurements were completed after a minimum of 24-hour cessation from exercise or physical activity.
Activity Energy Expenditure
Activity energy expenditure was assessed during each of the 5 phases throughout the entire basketball season. Each phase, subjects were outfitted with integrated heart rate and uniaxial accelerometer activity monitors for 4 consecutive days (Acti-Heart; CamNTech, Inc., Boerne, TX). Monitors were worn on the left side of the athlete's chest below their left breast. The monitors consisted of 2 locations that were attached to the chest using a standard electrocardiography electrode. According to manufacturer recommendations, one location was at the level of the xiphoid at the anterior midline or equidistant between but lower than the typical 12-lead electrode locations: V1 and V2. The other location was located laterally approximately at the anterior midline. Both electrode locations were adjusted to ensure the lead wire was parallel to the ground. Heart rate was continuously collected every minute for the entire movement period and was combined with accelerometry data to calculate activity energy expenditure. On most weeks, this included a game day, 2 practice days, and an off day. The Acti-Heart software computes activity energy expenditure, predicts resting metabolic rate using the equation of Schofield (28), and from these data, assigns a value for thermic effect of food that is fixed at 10 percent of the computed TDEE. A physical activity level (PAL) value was then calculated using the TDEE data divided by resting metabolic rate. Average total energy expenditure over the 4-day period was used to determine both TDEE and activity energy expenditure rates (presented as mean ± SD kcal·d−1). This information was also included and used in conjunction with dietary intake for the calculations for energy availability and energy balance. Previously published work by Assah et al. (1,4) in a large group of free-living adults revealed that the combined heart rate and activity monitors were valid in comparison with doubly labeled water.
Dietary Energy Intake
Dietary energy intake was assessed from 4-day diet logs completed during the same days that energy expenditure was monitored. Subjects were given food log packets that illustrated how to accurately record portion sizes of various foods and beverages consumed. Athletes logged all calorie-containing food and beverages for 4 consecutive days through Track—Calorie Counter (Nutritionix, New York, NY), an online software application downloaded on their smartphones. Subjects were instructed to log every eating occasion through the software at all feeding times. The average dietary intake over the 4-day period was used to assess energy intake through each phase of the season. Macronutrient intakes were also assessed through the application and expressed as both absolute and relative to body mass.
Sleep and Recovery Assessment
On the data gathering days when activity monitors were outfitted, and diet logs were recorded, each subject was required to complete a recovery questionnaire that was used to quantify their sleep and recovery. The questionnaire included average hours of sleep per night, how many meals per day were consumed, and their levels of restlessness, soreness, quality of sleep, and satisfaction of training were all assessed using a 100-mm anchored visual analog scale according to the methods of Price et al. (24). Soreness levels were quantified by how sore they felt over the 4 consecutive days of data collection. Sleep was defined as the interval between the point at which the athletes fell asleep to the time when they woke up again the next morning. Questions were open-ended, and responses were largely self-defined by the athletes, particularly regarding what constituted a meal or a snack. In this respect, meals were defined as a feeding where the athletes had consumed 2 or more food groups.
Statistical Analyses
All statistical analyses were completed using Microsoft Excel (Microsoft, Seattle, WA) and the Statistical Package for the Social Sciences, version 23 (IBM, Armonk, NY). A p value of 0.05 was used to make all statistical determination, and a trend was defined as a p value between 0.05 and 0.10. Data are presented as mean values ± SD. Changes between phases were determined using a factorial (within-subject) analysis of variance (ANOVA) with repeated measures on time. When a significant ANOVA was determine, individual pairwise comparisons were used to determine statistical significance.
Results
Of the 15 women on the basketball team, one was excluded by the IRB because of her role as a student investigator on the project, and 1 other transferred schools midseason, resulting in a sample size of 13. Descriptive information for all study subjects is provided in Table 1. A paired-samples t-test indicated the resting metabolic rate measurements made during phase I (1,857 ± 202 kcals·d−1) and phase III (1,801 ± 194 kcals·d−1) were highly correlated (r = 0.76, p = 0.002) and were not statistically different (p = 0.16). Self-reported hours of sleep (6.5 ± 0.72 hours per night) did not change (p = 0.16) throughout the study protocol.
Table 1. -
Subject demographics.
*
| Variable |
Mean ± SD
|
| Age (y) |
19.8 ± 1.3 |
| Height (cm) |
173.9 ± 13.6 |
| Body mass (kg) |
74.6 ± 9.1 |
| DEXA % fat |
27.1 ± 3.2 |
| DEXA fat mass (kg) |
20.4 ± 4.0 |
| DEXA fat-free mass (kg) |
52.8 ± 6.6 |
| Resting metabolic rate (kcals·d−1)† |
1,829 ± 185 |
| Resting metabolic rate normalized to body mass in kg (kcal·kg−1·d−1) |
24.7 ± 2.7 |
| V̇o
2 (ml·kg−1·min−1) |
53.9 ± 6.4 |
*DEXA = dual-energy x-ray absorptiometry.
†Reported resting metabolic rate value is the average of both measurements taken across the data collection period.
Visual Analog Scales
Anchored 100-mm visual analog scales were completed to assess feelings of restfulness, soreness, sleep quality, and training satisfaction. Feelings of restfulness significantly increased (phase I: 50.5 ± 13.0 mm, phase II: 49.9 ± 12.7 mm, phase III: 52.1 ± 13.4 mm, phase IV: 59.8 ± 12.7 mm, phase V: 65.4 ± 12.1 mm, p = 0.007) across the season. Phase IV was significantly greater than phases II (p = 0.047) and III (p = 0.020) but was not different from phase I and phase V. Phase V was significantly greater (p < 0.05) than phase I, II, and III but was not different from phase IV. Changes in soreness did not change (p > 0.05) across the season (phase I: 59.9 ± 20.9 mm, phase II: 60.2 ± 22.9 mm, phase III: 61.7 ± 17.5 mm, phase IV: 65.8 ± 16.8 mm, and phase V: 56.4 ± 23.8 mm). Assessments of sleep quality improved across the season (p = 0.001). Pairwise comparisons revealed that phase I sleep quality ratings (48.2 ± 17.5 mm) were significantly lower than phase IV (65.1 ± 15.4 mm, p = 0.003) and phase V (70.5 ± 11.4 mm, p < 0.001). Phase V sleep quality ratings were significantly greater than phase I (p < 0.001), II (p = 0.003), and III (p = 0.016). Ratings of training satisfaction tended to change across the season (phase I: 79.1 ± 9.5 mm, phase II 70.5 ± 12.4 mm, phase III: 67.8 ± 11.1 mm, phase IV: 76.7 ± 9.6 mm, phase V: 76.8 ± 11.9 mm, p = 0.069). Individual pairwise comparisons revealed that phase II scores (p = 0.045) and phase III (p = 0.005) were significantly less than phase IV.
Energy Expenditure
Using factorial ANOVA with repeated measures, TDEE tended to decrease throughout the season (p = 0.059). Individual groupwise comparisons revealed that phase III total energy expenditure levels were greater than phase IV (phase III: 2,850 ± 159 kcals·d−1 vs. phase IV: 2,674 ± 216 kcals·d−1; p = 0.001). Activity energy expenditure values did not change across the season (p = 0.17) (Table 2). Physical activity levels (TDEE/REE) in phase IV (1.52 ± 0.17, p = 0.001) were significantly less than during phase I (1.75 ± 0.27) (Table 2).
Table 2. -
Energy expenditure, energy availability, and energy balance across entire season.
*
| Intake/phase |
Phase I |
Phase II |
Phase III |
Phase IV |
Phase V |
| Total daily energy expenditure (kcals·d−1) |
3,065 ± 361 |
2,866 ± 363 |
2,850 ± 159 |
2,674 ± 216† |
2,806 ± 419 |
| Activity energy expenditure (kcals·d−1) |
1,196 ± 296 |
1,252 ± 774 |
1,028 ± 157 |
819 ± 160† |
969 ± 362 |
| Physical activity level (PAL) |
1.75 ± 0.27 |
1.63 ± 0.22 |
1.62 ± 0.15 |
1.52 ± 0.17† |
1.59 ± 0.23 |
| Energy availability (kJ·kg−1 FFM) |
199.3 ± 21.8 |
188.6 ± 49.3 |
186.5 ± 44.5 |
199.4 ± 19.7 |
193.3 ± 28.1 |
| Energy balance (kcals·d−1) |
−767 ± 426 |
−757 ± 720 |
−705 ± 642 |
−212 ± 466 |
−291 ± 551 |
*FFM = fat-free mass.
†Different from phase I.
Energy Balance and Energy Availability
Energy balance calculations were negative at all phases of data collection and did not statistically change from 1 part of the season to another (p = 0.64). Energy availability (kJ·kg−1 FFM) was calculated and did not change statistically (p = 0.81) (Table 2).
Dietary Intake
Absolute (2,425 ± 218 kcals·d−1, p = 0.72) and normalized (32.7 ± 3.7 kcals·kg−1·d−1, p = 0.79) daily caloric intake did not change across the season. Absolute (p = 0.29) and normalized (p = 0.13) daily carbohydrate intake did not change across the season. Using factorial ANOVA with repeated measures, absolute (p = 0.07) and normalized (p = 0.07) daily protein intake both tended to change across the season. Individual groupwise comparisons revealed that protein intake late in the competitive season (phase IV) was significantly lower (p = 0.02) than early in the competitive season (phase I). Absolute (100.9 ± 17.0 g·d−1, p = 0.10) and normalized (1.37 ± 0.25 g·kg−1·d−1, p = 0.32) daily fat intake did not change across the season. The average number of meals (3.7 ± 0.7 meals per day, p = 0.025) changed across the season from a high of 4.2 ± 1.0 meals per day during phase II to a low of 3.3 ± 0.65 meals per day during phase V. All dietary intake data can be found in Table 3.
Table 3. -
Dietary intake and meals consumed across entire season.
| Intake/phase |
Phase I |
Phase II |
Phase III |
Phase IV |
Phase V |
| Calories (kcal·d−1) |
2,506 ± 271 |
2,354 ± 533 |
2,326 ± 456 |
2,517 ± 334 |
2,422 ± 276 |
| Caloric intake (kcals·kg−1·d−1) |
33.7 ± 3.1 |
31.9 ± 7.8 |
31.5 ± 7.3 |
33.8 ± 3.7 |
32.7 ± 4.9 |
| Carbohydrates (g·d−1) |
282.4 ± 60.3 |
272.2 ± 73.2 |
244.8 ± 42.2 |
299.9 ± 36.4 |
263.2 ± 36.8 |
| Carbohydrates (g·kg−1·d−1) |
3.8 ± 0.7 |
3.7 ± 1.1 |
3.3 ± 0.7 |
4.0 ± 0.4* |
3.6 ± 0.7 |
| Protein (g·d−1) |
97.9 ± 18.8 |
87.3 ± 13.9 |
87.5 ± 17.0 |
78.0 ± 13.9† |
84.7 ± 16.3 |
| Protein (g·kg−1·d−1) |
1.31 ± 0.22 |
1.18 ± 0.19 |
1.19 ± 0.28 |
1.05 ± 0.19† |
1.15 ± 0.26 |
| Fat (g·d−1) |
113.0 ± 26.1 |
98.4 ± 27.1 |
112.7 ± 29.3 |
87.3 ± 18.8*† |
93.3 ± 28.5 |
| Fat (g·kg−1·d−1) |
1.50 ± 0.34 |
1.37 ± 0.38 |
1.51 ± 0.42 |
1.23 ± 0.31* |
1.23 ± 0.35 |
| Meals (meals per day) |
3.8 ± 0.9 |
4.2 ± 1.0† |
3.5 ± 0.9 |
3.6 ± 0.8 |
3.3 ± 0.6 |
*Different from phase III.
†Different from phase I.
Discussion
The primary findings of this study indicate that energy expenditure levels (total and activity) were highest early in the season and were maintained throughout the season with the exception of the period closest to the end of the competitive season (phase IV). Energy balance was negative across the entire season, and calorie, carbohydrate, and fat intake did not change while protein intake (Table 2) significantly decreased in phase IV. Energy availability did not change throughout the season while total energy expenditure, activity energy expenditure, and PALs experienced significant reductions in latter parts of the season when compared with early parts of the season.
Only recently have more reports begun to appear providing energy turnover data in team sports (5,9,19,29,30). Historically, it has been well documented that a variety of endurance athlete populations have a large negative energy balance throughout their competitive seasons (13), but this relationship within team-sports of varying intensities (i.e., stop and go sports) is known to a lesser degree. In agreement with our work, Briggs et al. (5) examined the changes in energy intake and energy expenditure in 10 male professional, adolescent soccer players during 1 week of competitive training (4 training days, 2 rest days, and 1 match day). A negative energy balance (−311 ± 397 kcals) was found across all days with the lowest energy balance levels being found on training and match days. Reed et al. (25) assessed the energy availability of Division I collegiate female soccer players during the preseason, midpoint, and postseason. As seen in this study, energy availability was lowest throughout the season when compared with preseason or postseason values.
Using doubly labeled water to assess total energy expenditure and a 4-compartment model for body composition, Silva et al. (29,30) reported that female athletes, when measured in preseason and postseason, were able to largely maintain energy balance and avoid changes in body composition. Unfortunately, we were not able to repeatedly measure body mass and body composition because of concerns by the coaching staff of preoccupation by the athletes with their results. These data would have allowed us to better outline if the reported negative energy balance levels were impacting body mass and body composition. Interestingly, Silva reported an average energy balance of only −14 kcals·d−1. When comparing the individual components of energy balance, it seems the source of discrepancy between our data lies with the reported daily caloric intake. When considering dietary intake data, values reported by Reed et al. (25) also exhibited a similar pattern of calorie intake as this study, whereby caloric intake was highest during the initial data collection period (phase I) and lowest during the 2 data collection periods that occurred during the season (Table 3). Furthermore, carbohydrate intake was successfully maintained across the season, but their reported daily intakes (3.3–4.0 g·kg−1·d−1) were lower than commonly recommended amounts for this athletic population (6,7,16). Protein intake during phase IV was significantly lower than values reported in phase I, and like carbohydrate, dietary protein intake failed to meet recommended daily amounts at any time point (14,16). To the authors' knowledge, no other study has assessed energy expenditure, energy intake, body composition, and resting metabolism to assess energy turnover in a competitive group of female team-sport athletes with this frequency or across an entire competitive season. It is important to consider that as with other published literature that used dietary recall approaches to assess energy intake (15,26,32), the potential for error in these measurements exists. However, previous results from Basiotis et al. (2) provide assurance that our assessment of energy and macronutrient intake was representative of the entire group of athletes in this study. Similarly, while other reports have indicated an overestimation of energy expenditure using the combined heart and activity monitor as used in this study (23), this work used the monitors for less than 1 week and only measured energy expenditure at 1 point while the monitors in this study were worn on 5 different occasions for 4 consecutive days.
In conclusion, a season-long investigation into changes in energy expenditure and energy intake consisting of 5 separate, identical measurement periods in a single cohort of collegiate female basketball players indicated that energy expenditure and energy intake rates were appropriately maintained through all parts of the season with the exception of the final months of the season.
Practical Applications
Results from this study should help to reinforce to sport coaches, strength and conditioning coaches, athletic trainers, athletes, and researchers that achieving an appropriate energy balance can be a challenge for many athletes. In addition, information from this study provides valuable insight into how well athletes are meeting general and specific nutrition recommendations across multiple points in a competitive season. This information can be used to bolster nutrition education to heighten awareness or to increase access to food to help offset potential dietary shortcomings.
Acknowledgments
The authors thank all the study subjects for their assistance with the study protocol. Special thanks are also extended to Mallory Eitel for her insight and support toward the completion of the project.
References
1. Assah FK, Ekelund U, Brage S, Wright A, Mbanya JC, Wareham NJ. Accuracy and validity of a combined heart rate and motion sensor for the measurement of free-living physical activity energy expenditure in adults in Cameroon. Int J Epidemiol 40: 112–120, 2011.
2. Basiotis PP, Welsh SO, Cronin FJ, Kelsay JL, Mertz W. Number of days of food intake records required to estimate individual and group nutrient intakes with defined confidence. J Nutr 117: 1638–1641, 1987.
3. Bolonchuk WW, Lukaski HC, Siders WA. The structural, functional, and nutritional adaptation of college basketball players over a season. J Sports Med Phys Fitness 31: 165–172, 1991.
4. Brage S, Brage N, Franks PW, Ekelund U, Wareham NJ. Reliability and validity of the combined heart rate and movement sensor Actiheart. Eur J Clin Nutr 59: 561–570, 2005.
5. Briggs MA, Cockburn E, Rumbold PL, Rae G, Stevenson EJ, Russell M. Assessment of energy intake and energy expenditure of male adolescent academy-level soccer players during a competitive week. Nutrients 7: 8392–8401, 2015.
6. Burke LM, Hawley JA, Wong SH, Jeukendrup AE. Carbohydrates for training and competition. J Sports Sci 29(Suppl 1): S17–S27, 2011.
7. Burke LM, Loucks AB, Broad N. Energy and carbohydrate for training and recovery. J Sports Sci 24: 675–685, 2006.
8. Camic CL, Kovacs AJ, Enquist EA, Vandusseldorp TA, Hill EC, Calantoni AM, et al. An electromyographic-based test for estimating neuromuscular fatigue during incremental treadmill running. Physiol Meas 35: 2401–2413, 2014.
9. Devlin BL, Kingsley M, Leveritt MD, Belski R. Seasonal changes in soccer players' body composition and dietary intake practices. J Strength Cond Res 31: 3319–3326, 2017.
10. Drenowatz C, Eisenmann JC, Carlson JJ, Pfeiffer KA, Pivarnik JM. Energy expenditure and dietary intake during high-volume and low-volume training periods among male endurance athletes. Appl Physiol Nutr Metab 37: 199–205, 2012.
11. Fagerberg P. Negative consequences of low energy availability in natural male bodybuilding: A review. Int J Sport Nutr Exerc Metab 28: 385–402, 2018.
12. Gonzalez AM, Hoffman JR, Scallin-Perez JR, Stout JR, Fragala MS. Performance changes in National Collegiate Athletic Association Division I women basketball players during a competitive season: Starters vs. nonstarters. J Strength Cond Res 26: 3197–3203, 2012.
13. Heydenreich J, Kayser B, Schutz Y, Melzer K. Total energy expenditure, energy intake, and body composition in endurance athletes across the training season: A systematic review. Sports Med Open 3: 8, 2017.
14. Jager R, Kerksick CM, Campbell BI, Cribb PJ, Wells SD, Skwiat TM, et al. International society of sports nutrition position stand: Protein and exercise. J Int Soc Sports Nutr 14: 20, 2017.
15. Jeacocke NA, Burke LM. Methods to standardize dietary intake before performance testing. Int J Sport Nutr Exerc Metab 20: 87–103, 2010.
16. Kerksick CM, Arent S, Schoenfeld BJ, Stout JR, Campbell B, Wilborn CD, et al. International society of sports nutrition position stand: Nutrient timing. J Int Soc Sports Nutr 14: 33, 2017.
17. Logue D, Madigan SM, Delahunt E, Heinen M, Mc Donnell SJ, Corish CA. Low energy availability in athletes: A review of prevalence, dietary patterns, physiological health, and sports performance. Sports Med 48: 73–96, 2018.
18. Loucks AB. Energy balance and body composition in sports and exercise. J Sports Sci 22: 1–14, 2004.
19. Mara JK, Thompson KG, Pumpa KL. Assessing the energy expenditure of elite female soccer players: A preliminary study. J Strength Cond Res 29: 2780–2786, 2015.
20. Mielgo-Ayuso J, Zourdos MC, Calleja-Gonzalez J, Urdampilleta A, Ostojic SM. Dietary intake habits and controlled training on body composition and strength in elite female volleyball players during the season. Appl Physiol Nutr Metab 40: 827–834, 2015.
21. Mountjoy M, Sundgot-Borgen J, Burke L, Carter S, Constantini N, Lebrun C, et al. The IOC consensus statement: Beyond the female athlete triad–relative energy deficiency in sport (RED-S). Br J Sports Med 48: 491–497, 2014.
22. Nana A, Slater GJ, Hopkins WG, Halson SL, Martin DT, West NP, et al. Importance of standardized DXA protocol for assessing physique changes in athletes. Int J Sport Nutr Exerc Metab 26: 259–267, 2016.
23. Nichols JF, Aralis H, Merino SG, Barrack MT, Stalker-Fader L, Rauh MJ. Utility of the actiheart accelerometer for estimating exercise energy expenditure in female adolescent runners. Int J Sport Nutr Exerc Metab 20: 487–495, 2010.
24. Price DD, Mcgrath PA, Rafii A, Buckingham B. The validation of visual analogue scales as ratio scale measures for chronic and experimental pain. Pain 17: 45–56, 1983.
25. Reed JL, De Souza MJ, Williams NI. Changes in energy availability across the season in Division I female soccer players. J Sports Sci 31: 314–324, 2013.
26. Scagliusi FB, Ferriolli E, Pfrimer K, Laureano C, Cunha CS, Gualano B, et al. Underreporting of energy intake in Brazilian women varies according to dietary assessment: A cross-sectional study using doubly labeled water. J Am Diet Assoc 108: 2031–2040, 2008.
27. Schaal K, Tiollier E, Le Meur Y, Casazza G, Hausswirth C. Elite synchronized swimmers display decreased energy availability during intensified training. Scand J Med Sci Sports 27: 925–934, 2017.
28. Schofield WN. Predicting basal metabolic rate, new standards and review of previous work. Hum Nutr Clin Nutr 39(Suppl 1): 5–41, 1985.
29. Silva AM, Matias CN, Santos DA, Thomas D, Bosy-Westphal A, Mu LM, et al. Compensatory changes in energy balance regulation over one athletic season. Med Sci Sports Exerc 49: 1229–1235, 2017.
30. Silva AM, Matias CN, Santos DA, Thomas D, Bosy-Westphal A, Muller MJ, et al. Energy balance over one athletic season. Med Sci Sports Exerc 49: 1724–1733, 2017.
31. Silva AM, Santos DA, Matias CN, Rocha PM, Petroski EL, Minderico CS, et al. Changes in regional body composition explain increases in energy expenditure in elite junior basketball players over the season. Eur J Appl Physiol 112: 2727–2737, 2012.
32. Stubbs RJ, O'reilly LM, Whybrow S, Fuller Z, Johnstone AM, Livingstone MB, et al. Measuring the difference between actual and reported food intakes in the context of energy balance under laboratory conditions. Br J Nutr 111: 2032–2043, 2014.
33. Tanaka H, Monahan KD, Seals DR. Age-predicted maximal heart rate revisited. J Am Coll Cardiol 37: 153–156, 2001.
