Total Energy Expenditure, Energy Intake, and Body Composition in Endurance Athletes Across the Training Season: A Systematic Review

Total energy expenditure (TEE) is composed of the energy costs of the processes essential for life (basal metabolic rate (BMR), 60–80% of TEE), of the energy expended in order to digest, absorb, and convert food (diet-induced thermogenesis, ~10%), and the energy expended during physical activities (activity energy expenditure, ~15–30%) [1, 2]. Elite endurance athletes are characterized by high fluctuations of TEE, mainly due to the variability of the energy expended during sporting activities. Among elite senior endurance athletes, training loads from 500 h/year [3, 4] up to 1000 h/year [5‐7] have been reported, depending on the specific muscular loading characteristic of the sport. During heavy sustained exercise (e.g., during the Tour de France), TEE can be as high as fivefold the BMR over several weeks [8]. On the other hand, during recovery days, pre-competition tapers, or during the off-season, the energy expended in activities is far less. Therefore, TEE is expected to be much lower and may even reach levels comparable to that of sedentary behavior.

An appropriate energy intake supports optimal body function, determines the capacity for intake of macronutrients and micronutrients, and assists in manipulating body composition in athletes [9]. It is a challenge for each endurance athlete to appropriately match energy intake and TEE in order to achieve energy balance and thus, weight stability, both on a micro level (i.e., over 1 day or several days) and through the training and competitive season. Furthermore, endurance athletes in general strive for a low body mass and/or body fat level for various advantages in their sports, specifically during the competition season [10]. This allows runners and cyclists to reach greater economy of movement and better thermoregulatory capacity from a favorable ratio of weight to surface area and less insulation from subcutaneous fat tissue. Elite endurance athletes are therefore characterized by low body mass and body fat content. For example, in elite Kenyan endurance runners, the body fat percentage was 7.1% [11], which is only marginally above the recommended 5% minimum for males [12]. In the same athletes, body mass index (BMI) was 18.3 kg/m² [11], which is generally classified as being underweight [13]. However, these athletes were in peak physical conditions as the investigations were undertaken and a low body fat percentage and body weight might be an advantage for competition. Achieving a negative energy balance and a concomitant loss of body and fat masses in preparation for competition can be accomplished in phases with high daily TEE solely by the reduction of energy intake, since any further training load increases could cause overtraining [12]. Therefore, the nutritional goals and requirements of endurance athletes are not static over the training year. Since endurance athletes undertake a periodized training program and follow periodized body composition goals, the nutritional support also needs to be periodized [9].

Usually, the annual training schedule of an elite endurance athlete is divided into distinct phases, each with very specific objectives. This is necessary to maximize physiological adaptations for improved performance, usually scheduled to peak around the main competitions of the year [14]. The principle of training periodization was first introduced in the 1960s by the Soviet trainer Leo Matveyev [15] and has not fundamentally changed since then [14]. The basis of this model is to prepare the athlete for one or more major competitions during the year by separating the training into the following three main phases (macrocycles): preparatory, competitive, and transition phases [15]. An example for a “one-peak annual plan” for a runner is shown in Fig. 1. The preparatory phase is characterized by predominantly high-volume training at moderate intensities, which improves endurance capacity and provides a more efficient use of fuel substrates. During the late preparatory phase, training volume is reduced while intensity is gradually increased. The goal of this phase is to reach peak performance and to transfer the training effects into the competitive phase, where exercise intensity is the highest. In the week before an important competition, volume and intensity are typically decreased (taper phase) to allow the body to optimally recover for competition. The days and weeks after a main competition are characterized by low-intensity and low-volume training, with goals to induce regeneration and to prepare the athlete mentally and physically for the next training cycle (transition phase) [14, 16].

Although the concept of training periodization in elite endurance sports has been established for a long time, the coupling of periodized training with nutrition and body composition has gained scientific awareness only recently [17]. Stellingwerff’s group was one of the first to publish periodized nutrition guidelines for middle-distance athletes [17], they then expanded these recommendations for a multitude of power sports [18]. Nowadays, there are guidelines for carbohydrate, protein, and fat intake during training and competition phases, not exclusively focusing on endurance sports [19‐21]. Meanwhile, for endurance athletes, sport-specific dietary intake recommendations were developed only for a few endurance disciplines (e.g., swimmers [22‐25], distance runners [26], marathon/triathlon/road cycling [27]). But it remains unclear whether endurance athletes are actually following these nutrient guidelines across all seasonal training phases.

The validity of either body composition, energy intake, or TEE-determination in athletes strongly depends on the methods used. The measurement of body composition in general is prone to error. It has been shown that acute food or fluid ingestion [28], subject positioning [29], previous physical activity [30], and hydration status [31] have an impact on reliability of body composition measurement. Since endurance athletes often train several times per day, it might be difficult to assure best conditions for body composition assessment. According to a recent methodology review performed by Nana et al., only few of the studies, where body composition of athletes was measured with dual X-ray absorptiometry (DXA), provided details about their subject and device standardization [30]. However, other methods like skinfold measurements require highly experienced investigators [32] and strongly depend on the number of measurement sites and the formula used to calculate the percentage of body fat [33]. Therefore, it is important to report standardization protocols in order to evaluate the quality of data assessment. One main issue in assessing energy intake in athletes is the magnitude of under-reporting, which can amount to 10–45% of TEE [34]. It was shown that the magnitude of under-reporting increases as energy requirements increase [34]. Since endurance athletes are often characterized by high TEE, we must assume that these athletes are very prone to a high percentage of under-reporting. For determination of TEE objective methods such as doubly labelled water (DLW) or heart frequency measurements are available. However, in many studies subjective methods such as activity records and activity questionnaires are used in order to assess the activity level and TEE of subjects. These methods estimate TEE or activity level and their validity strongly depends on the breadth of the activity dimensions analyzed.

There exist some longitudinal studies that have assessed fluctuations in body composition, dietary intake, and/or TEE of endurance athletes across the training seasons [35‐52], but no systematic reviews have been performed. Therefore, the purpose of this study was to (1) systematically analyze TEE, energy intake, and body composition in highly trained athletes of various endurance disciplines and of both sexes with focusing on objective assessment methods and (2) analyze fluctuations in these parameters across the training season. We hypothesized that endurance athletes show large fluctuations of TEE during different seasonal training phases due to differing exercise loads, and concomitant alterations in energy intake and body composition.

The review protocol was developed according to the Meta-analysis of Observational Studies in Epidemiology Guidelines for meta-analyses and systematic reviews of observational studies [53].

In this systematic review, we examined fluctuations in TEE, energy intake, and/or body composition in endurance athletes across the training season. We found that some, but not all, of the investigated outcomes depended on the time point of data assessment during seasonal training. TEE was highest during the competition phase and higher than energy intake in all seasonal training phases. Alterations in TEE did not lead to adaptations of energy intake in females, whereas in males, a higher absolute energy intake during the competition phase was observed. The finding that male endurance athletes demonstrated the highest fat mass values during the competition phase and the lowest FFM during the transition phase seems to be an anomaly from the pooling of data.

Our systematic search initially yielded many studies where TEE, energy intake, or body composition in endurance athletes were investigated. Only a few (2%) reported the time point of data collection with regard to the training season and could thus be included in this review. This is unfortunate since our analysis clearly illustrates how training volume and related TEE vary importantly with seasonal training phases. Specifically and expectedly, both absolute and relative TEEs were significantly higher during the competition phase compared to the preparation phase. Interestingly, these differences were only partly in agreement with alterations in energy intake and/or body composition of endurance athletes.

During the transition phase, limited data for TEE and energy intake of endurance athletes was available. Only for body composition, it was possible to compare with other seasonal training phases, although the number of study estimates and therefore, explanatory power, was weak. Future research on elite athletes should focus on the effects of a sudden stop or reduction in TEE on body composition (e.g., because of injury). There exist only a few studies (with conflicting results) where this question has been examined. Ormsbee and Arciero investigated the effects of 5 weeks of detraining on body composition and RMR in eight male and female swimmers [65]. RMR decreased, whereas fat mass and body weight increased with detraining. In contrast, LaForgia et al. showed that after 3 weeks of detraining, no differences in RMR and percentage of fat mass occurred in male endurance athletes [38]. Unfortunately, energy intake was not reported in either of these studies. Thus, it remains unclear when, whether, and to what extent the body adapts (through changes in energy intake and/or body composition) for the decrease in TEE caused by detraining.

Our analysis highlights an important apparent negative energy balance in endurance athletes, both in the preparation and competition phases, when separately examining the energy balance in articles where both energy intake and TEE were assessed (N = 11). Negative energy balance was reported during the preparation phase in male [66, 67] and female [67] cross-country skiers, male [11] and female [68] runners, and female lightweight rowers [69] and swimmers [70], and amounted to a mean of 304 kcal/day (4.7% of TEE) for males and 1145 kcal/day (27.8%) for females. During the competition phase, a negative energy balance was reported in male cyclists and triathletes [60], male [63] and female [63, 71] runners, and male cyclists [61, 62], averaging 2177 kcal/day (32.5%) for male and 1252 kcal/day (47.9%) for female endurance athletes. The most obvious explanation for these energy deficits is likely the classical issue of under-reporting energy intake through self-assessment in human studies. A review of nine studies using DLW to validate self-reported energy intake in athletes revealed that under-reporting can amount to 10–45% of TEE [34]. Since under-reporting increases in magnitude as energy requirements increase [34], we must assume that under-reporting in the present study estimates was more important during the competition phase. Even when 45% was added to the energy intake of all athletes included in our review, there still remained a negative energy balance of 118 kcal (2.7% of TEE) in the preparation and 5293 kcal (53.6%) in the competition phase. Another explanation for the negative energy balance might be the low accuracy and precision of methods used to estimate energy intake in athletes in the articles included in our review. For example, mostly dietary records with a mean observation time of 4.7 ± 4.1 days were used. According to Magkos and Yannakoulia, for athletes, a 3–7-day diet-monitoring period would be enough for reasonably accurate and precise estimations of habitual energy and macronutrient consumption [34]. However, other methods like FFQs and dietary recalls were also used for energy intake estimations. These methods are both memory-dependent and show lower accuracy and precision than prospective methods like dietary records [72]. However, even when only articles were considered where energy intake was assessed by the use of dietary records, the error remained high (2.5% of TEE during the preparation phase and 54.9% during the competition phase). Finally, the high negative energy balance during the competition phase may also be explained by the fact that, apart from one study, all included studies investigated the TEE during the days with actual competition and not during habitual training days in the competition phase. Thus, it is likely that the TEE during this phase was over-estimated. During the preparation phase, a negative energy balance leading to increased energy store utilization might be desirable by coaches and athletes to reach a sport-specific body composition, but during the competition phase, body composition should not be modified anymore since it is typically already at its optimum. There was one study in which dietary intake was strictly controlled since the subjects were in confinement. Brouns et al. simulated a Tour de France race in a metabolic chamber and calculated the daily energy balance from the energy expended and energy intake as calculated from daily food and fluid consumption [73]. They found a positive energy balance during active rest days whereas during the exercise days, a significant negative energy balance was observed. The authors concluded that if prolonged intensive cycling increases energy expenditure to levels above a certain threshold (probably around 20 MJ or 4780 kcal), athletes are unable to consume enough conventional food to provide adequate energy to compensate for the increased energy expenditure. The authors of a recent review addressing the criticisms regarding the value of self-reported dietary intake data reasoned that these should not be used as a measure of energy intake [74]. Our analysis supports this statement since, for athletes, relative energy deficits amounted up to 48% of TEE in female athletes and 33% in male athletes during the competition phase. Thus, there is an urgent need for better methods of dietary intake quantification, such as dietary biomarkers and automated image analysis of food and drink consumption [74]. The classical concept of energy balance, defined as dietary energy intake minus TEE, has been criticized, since according to this definition energy balance is the amount of dietary energy added to or lost from the body’s energy stores after the body’s physiological systems have done their work for the day [75]. Thus, energy balance is an output from those systems. In contrast, energy availability, defined as the dietary energy intake minus the energy expended during exercise, is an input to the body’s physiological systems, since energy availability is the amount of dietary energy remaining for all other metabolic processes [75]. Endurance athletes, especially female athletes, show low energy availability (<30 kcal/kg FFM/day) [76] and increased risk for changes of the endocrine system affecting energy and bone metabolism, as well as in the cardiovascular and reproductive systems [77]. In healthy young adults, energy balance = 0 kcal/day when energy availability = 45 kcal/kg FFM/day [75]. Since the results of the present study indicate a high negative energy balance in endurance athletes, we must assume that the athletes also demonstrate low energy availability. However, due to the limited data, it was not possible to account for other clinical markers (e.g., bone mineral density), menstrual status, or prevalence of eating disorders in the athletes. We recommend that energy balance-related studies in endurance athletes should also assess and report clinical markers, such as bone mineral density and menstrual status, in order to assess the clinical consequences of the mismatch of TEE and energy intake.

The aggregate analysis yielded a surprising finding. In male endurance athletes, the absolute and relative fat mass was highest during the competition phase. In contrast, during the transition phase, FFM was lowest, which goes along with our expectations with a decrease in exercise volume and intensity. For the female athletes, we did not find these fluctuations in body composition, except for a higher body fat content during the preparation phase compared to the transition phase. We believe that these findings are due to the paucity of data and to the fact that the number and type of athletes varied between seasonal training phases. Indeed, when separately analyzing the few studies where body mass and composition were assessed during both the preparation and competition phases (N = 5), both male and female endurance athletes showed a significantly lower percentage of body fat and higher FFM during the competition phase. Further studies with longitudinal assessments of body composition are required to support these findings. However, in only 5.7% of the studies, where body composition was assessed, satisfactory details about standardization were provided. According to Nana et al., studies involving DXA scans of body composition should report details of the DXA machine and software, subject presentation and positioning protocols, and analysis protocols [30]. It has been shown that the use of a non-standardized protocol increased the variability for total and fat-free soft tissue mass compared to a standard protocol, which might include a loss in ability to detect an effect of an intervention that might have relevance for sports performance [78]. The use of non-standardized protocols and the concomitant higher variability might explain some of the unexpected findings of body composition changes in athletes of the present study.

In male endurance athletes, absolute energy intake was higher during the competition phase compared to the preparation phase. The relative energy intake was not different, which can be explained by the apparent significant increase of body mass during the competition phase, and is likely an artifact of the aggregation of data from various studies. In female athletes, neither absolute nor relative energy intake was different between seasonal phases. When focusing on longitudinal studies that assessed energy intake during different training seasons in the same cohort, there was a tendency for male athletes to show greater fluctuations in energy intake. In female cross-country skiers, the energy intake was higher during the preparation phase [50], whereas in female runners and swimmers, the energy intake was higher during the competition phase [47]. However, summing up both studies, no significant differences between training season phases were found. In contrast, male endurance athletes showed a significantly higher energy intake during the competition phase, as seen in male runners [44], cross-country skiers [50], swimmers [43], and triathletes [49]. Although some of the included studies showed greater energy intake in male endurance athletes during the preparation phase (cyclists [46, 48], swimmers [43]), the power of these studies was too low to change the results. However, since energy intake varies in male endurance athletes depending on the training season phase, it indeed seems appropriate to adapt dietary recommendations according to the different training season phases, as proposed by Stellingwerff et al. [17, 18].

Our analysis highlights the important seasonal fluctuations in TEE, energy intake, and body composition in male and female endurance athletes across the training season. Therefore, dietary intake recommendations should take into consideration other factors including the actual training load, TEE, and body composition goals of the athlete. The present review supports the statement of the current position stand of the American College of Sports Medicine (ACSM) that energy and nutrient requirements are not static and that periodized dietary recommendations should be developed [9]. Importantly, our analysis again shows the uselessness of self-reported dietary intake, a well-known limitation to energy balance studies, in endurance athletes. The important underreporting suggested by our analysis again raises the question of whether self-reported energy intake data should be used for the determination of energy intake and illustrates the need for more valid and applicable energy intake assessment methods in free-living humans [74]. Since we observed a lack of data during the transition phase, future research should focus on the assessment of TEE, energy intake, and body composition on a reduction in training intensity and volume, such as at the end of the competitive season. In addition, future studies dealing with energy balance and nutrient intake in elite endurance athletes should always mention the time point of data assessments (e.g., seasonal training phase).