Over a season, dynamic changes in body weight (BW) and its composition, fat mass (FM) and fat-free mass (FFM) are expected to occur but the magnitude of these values may vary widely according to demands of sports-specific competitions. Increases in BW, mainly through FFM gains, are relevant for sports, such as rugby, football, or hockey, where resulting strength gains may offer a protective advantage. In turn, athletes involved in weight-sensitive sports, such as gravitational sports (e.g., triathletes, runners, etc.), weight category sports (e.g., Judo, wrestling), and aesthetic sports (1), have a competitive advantage if weight is lost, specifically by reducing FM.
However, for sports professionals responsible for prescribing nutrition and exercise regimens over a season, a key question is still to be addressed: how do FM and FFM respond to these dynamic changes in BW in an athletic population? In the past century, medical scientists, using nonathletic lean and overweight/obese populations, provide us with conceptual models to understand how the energy partition ratio respond to changes in energy balance, specifically after diet-induced BW loss.
Wishnofsky (24) viewed diet-related BW changes as deriving solely from adipose tissue consisting of 87% of FM. Webster (23) theorized that diet-induced BW loss would consist of a rounded proportion of 75% as FM and 25% as FFM, also known as the Quarter FFM Rule. A key issue of both models is the prediction of a constant energy density of weight loss (Wishnofsky, ~7700 kcal·kg−1; and Webster, ∼7400 kcal·kg−1). Instead of a static view of the energy partition ratio, a later approach considered a FFM–FM relationship, known as the Forbes Curve (4). Using a sample with a large body composition variability (from lean to obese subjects), Forbes Curve predicted that (i) individuals with greater FM have higher FFM identified by a nonlinear relationship, (ii) changes in BW and FM are accompanied by changes in FFM described by the nonlinear relationship, and (iii) FFM loss or gain is larger when baseline FM is smaller.
Although diet-induced interventions in normal populations have been validated as following the Forbes Curve, it has never been tested whether changes in the energy partition ratio of a highly active population, varying widely according to sports-specific demands, follow the earlier established rules or Forbes predictions. This is particularly important for athletes that restrict energy intake for achieving a competitive advantage, typically observed in weight-sensitive sports (18). In fact, concerns about excessive leanness and other body composition disturbances have been raised among sports professionals. A position statement under the auspices of the IOC Medical Commission has recognized these concerns and recommends the use of dual x-ray absorptiometry (DXA) to monitor changes in body composition after weight loss interventions in athletes (18).
Still, DXA is not the state of the art method for body composition assessment at the molecular level (21). Instead, the four-compartment model (4C-model) is considered the criterion method for assessing FM, as the major FFM molecular components are estimated, that is total body water, mineral and protein, and few assumptions are made for FM estimation (12). Nonetheless their advantages, limited studies have used 4C-models to determine body composition changes in athletic populations (15,16).
In experimental exercise interventions that do not seek to alter diet, FM reductions were associated to modest changes in FFM (2,8,14). Full comprehension of the dynamic changes in BW composition over a season, in particular by using a 4C-model to determine FM and FFM, is required to help developing optimal nutrition and exercise regimens that enhance athletic performance without compromising health. Therefore, the aim of this study was to analyze the longitudinal changes in the composition of BW change in highly trained athletes over one season and to analyze if these dynamic changes follow the theoretical driven rules or predictions derived from nonathletic lean or overweight/obese populations.
A total of 70 athletes, 42 men and 28 women participating in national and international championships were recruited from the following sports: basketball (n = 24), handball (n = 12), volleyball (n = 15), triathlon (n = 7), and swimming (n = 12).
Subject inclusion criteria were: 1) >10 h training per week, 2) negative test outcomes for performance enhancing drugs, and 3) not taking any medications or dietary supplements. Female athletes reporting a regular menstrual cycle were included in the study and were assessed during the luteal phase. All participants were informed about the possible risks of the investigation before giving their written informed consent to participate. The study was approved by the Ethics Committee of the Faculty of Human Kinetics, University of Lisbon and conducted in accordance with the declaration of Helsinki (25).
This was an observational study with a follow-up over one season ranging from 5 to 10 months, depending on each sport’s competitive timetable (~5 months for volleyball, ~8 months for basketball, ~7 months for handball, ~7 months for swimmers, and ~10 months for triathletes). The first period of evaluation was assessed in the beginning of the season, whereas the second period corresponded to the competitive phase, which occurred just before the main stage of the competition.
Basketball, handball, and volleyball player’s training regimens consisted of five sessions per week with a total of 2 h each, divided in technical-tactical training. Twice a week athletes performed resistance training for nearly 1 h before the sport practices. In general, resistance training progressed from general conditioning (weeks 1–4), to hypertrophy (weeks 5–8), then to maximal strength and power (weeks 9–16), and concluded with specific strength training (weeks 17–32).
Swimmer’s training regimens consisted of five sessions per week with a total of 150 min each, from which 90 min of general preparation training was carried out in the water (with a total volume of ~23.4 km·wk−1) and an additional 60 min of general strength training performed out of the water.
Triathlete’s training regimens consisted of sessions divided in swimming, cycling, running, and general resistance training. In the beginning of the preparatory period the training regimens consisted of five swimming sessions in the pool (volume, 19 km·wk−1), three sessions of cycling (volume 150 km·wk−1), and five sessions of running (volume, 60 km·wk−1). In the main stage of the competition the training regimens consisted of five swimming sessions in the pool and one in the ocean (pool + ocean volume, 30 km·wk−1), four sessions of cycling (volume 240 km·wk−1) and six sessions of running (volume, 60 km·wk−1). Additionally, during both assessments athletes were performing three sessions per week of general resistance training (volume, 90 min·wk−1).
Competitions were relegated to weekends during the competitive phase with at least one game per weekend integrated in the National Championships.
During both measurement periods, simultaneous assessments of body composition were carried out in the morning following a 12-h fast with no caffeine and alcohol during the preceding 24-h, and no vigorous exercise within 15 h.
BW and height were measured to the nearest 0.01 kg and 0.1 cm. Athletes were empirically divided into those who lost (n = 20) or gained >1.5% BW (n = 50).
FM and FFM
A 4C-model was used to assess body composition, calculated after using the soft tissue mineral (Ms) component obtained as Ms = 0.0129TBW (22). The 4C-model is described as follows:
where BV is body volume (L), TBW is total body water (kg), Mo is bone mineral (kg), and BM is body mass (kg).
Propagation of measurement error
The error associated with measurement of FM from the 4C-model can be estimated by assuming an average body composition of the whole sample (BV, TBW, Mo, and BW) and measurement precision of each method, as described elsewhere (17). The error associated with measurement of FM from the 4C-model is 0.7 kg.
TBW was assessed through deuterium dilution by a Hydra stable isotope ratio mass spectrometer (PDZ, Europa Scientific, Cheshire, UK). Procedures are described elsewhere (17) with a coefficient of variation (CV) of 0.3%. Bone mineral content was determined by dual-energy x-ray absorptiometry (Explorer-W, fanbeam densitometer, software QDR for windows version 12.4; Hologic, Waltham, MA), and converted to Mo by multiplying it by 1.0436 (11) with a CV of 1.3% (17). DXA was also used as an independent method to determine FM and FFM.
Air displacement plethysmography (COSMED, Rome, Italy) was used to assess BV with a CV of 0.4%, as described elsewhere (17).
Protein was calculated as BW minus FM (from the 4C-model), TBW, Mo, and Ms contents of FFM.
Energy partition ratio
According to Hall (8), the majority of the mathematical models of human energy regulation assume that a change in weight is uniquely determined by an imbalance between energy intake and energy expended to perform physical work and maintain life (8). The theoretical foundation of this energy balance concept is the first law of thermodynamics (8). Energy balance models can be mathematically termed as:
where the left side of the equation is the rate of change of body energy, with BW being the BW and ρ being an energy density converting between units of metabolizable energy and mass (8). The right side is the energy imbalance between the body’s energy intake rate, EI, and energy expenditure rate, EE. Any of the terms in the above equation can depend on time, t, as well as other parameters. If energy balance is negative, BW is reduced and EE is larger in magnitude than EI. Likewise, if energy balance is positive, BW increase and EE is smaller in magnitude than EI. EB represents the average energy deficit or surplus and can be calculated from the changed body energy stores from the beginning to the competitive-phase of the season. Using the established energy densities of 1.0 kcal·g−1 for FFM (3) and 9.5 kcal·g−1 for FM (13), we applied the formula below to quantify the energy density (ρ) of weight change:
where ΔFFM and ΔFM represent the change in g of FFM and FM from beginning to end of the follow-up. As described by Hall (8) the energy partition ratio ranges between 0 and 1 and determines the proportion of an energy imbalance directed to and from FFM versus FM.
Food records were also collected to characterize macronutrient composition of the diet in both moments. Dietary records were analyzed using a software package (Food Processor SQL; ESHA Research, Salem, OR).
Statistical analysis was performed using PASW Statistics for Windows version 22.0, 2013 (SPSS Inc., an IBM Company, Chicago, IL). Descriptive statistics including means ± SD were calculated for all the variables. Comparison of means were performed using one sample t-tests (mean difference from zero or a given value), paired sample t-tests (mean difference between moments), independent sample t-tests (mean difference between gender and weight categories), and one-way ANOVA (mean differences between sports). ANCOVA was used to test differences, adjusting for gender, sport type, and age.
Multiple regression analysis was used to test the validity of Forbes’ Curve and to determine the association between the energy partition ratio with initial body stores.
This study is able to reject the null hypothesis that a true difference is zero with a probability (power) of 80% and a Type I error probability of 0.05, if the mean energy partition ratio for FFM between the two weight change groups are 0.85 vs 0.06 (difference of 0.79) with a standard deviation of 0.90 vs 0.91, respectively, for those who lost or gain >1.5% BW.
For all tests, statistical significance was set at P < 0.05.
Summary of participants’ characteristics
For the whole-sample BW and FFM significantly increased by 1.8% ± 3.4% and 2.5% ± 4.1%, whereas FM reduced by 1.6% ± 18.7%, as observed in Table 1. No mean changes in energy, protein, carbohydrates, and fat intake from the dietary records completed at both time points were observed over the season.
Energy partition ratio by weight changes
Table 2 depicts body composition for the athletes that lost and gained >1.5% BW, ranging from −7.0% to −1.5% and from 1.5% to 8.2%, correspondingly.
A mean FM reduction of −1.7 kg ranging from −5.4 to 2.0 kg was observed in athletes that lost >1.5% BW with a large variability in individual FFM changes (−4.7 to 2.8 kg) but nonsignificant mean changes in FFM of −0.7 kg, indicating that some athletes were able to reduce FM while increasing FFM. In those who gained >1.5% BW, FFM significantly increased by 2.3 kg ranging from −3.6 to 8.0 kg with nonsignificant FM gains of 0.4 kg but again with a large individual variability (−6.3 to 5.2 kg).
A total of 9 women and 11 men lost more than 1.5% BW whereas 19 women and 31 men gained more than 1.5% BW. Mean differences between sex were found (P < 0.05) for FM (M1 and M2), FFM (M1 and M2), mineral (M1 and M2), protein (M1 and M2), water (M1 and M2) for those who lost or gain more than 1.5% BW. All the values were significantly lower in women compared with men, except for FM. Differences between moments did not differ by sex (P > 0.05) for those who lost more than 1.5% BW, whereas changes in FM, FFM, mineral, protein, and water differed in those who gained more than 1.5% (P < 0.05). Women presented a positive change in FM, a lower FFM, mineral, and water increase compared to men but a higher increase in protein compared to their counterparts.
The mean contribution of the changes in FM and FFM to BW changes, adjusted for sex, age, and sports, is displayed in the right panel of Figure 2 for those that lost or gain >1.5% BW, whereas the mean energy density of the weight change (kcal·kg−1) is showed in the left panel. Athletes who lost >1.5% BW were able to maximize the use of FM by ~90%, whereas those who gained >1.5% BW showed a contribution of ~95% directed to FFM (Fig. 1).
Actual versus predicted change in FFM did not follow the line of identity (Fig. 2). While there was an association between the two (R2 = 0.41 for those who lost BW and R2 = 0.65 for those who gain BW), the graph demonstrates that Forbes Curve is invalid for the case of predicting FFM change over one athletic season.
Energy partition ratio and baseline energy stores
Using DXA as an independent body composition method, baseline FM and FFM were tested as significant predictors of the energy partition ratio. No significant association was observed between initial FM and the energy partition ratio (P = 0.058) but initial FFM was significantly related (P = 0.044). After adjusting for sex, age, and sport type this later association was no longer significant (P = 0.488).
The novel finding from this investigation is that theoretical driven rules and predictions derived from nonathletic lean or overweight/obese populations are not applicable for dynamic body composition changes at the molecular level in highly trained athletes. Energy partition ratio involving FFM and FM changes was found to deviate from commonly used rules (2,3,4).
Energy partition ratio
Overall, our findings indicate that FFM significantly increased by approximately 4% in those athletes that changed their BW by more than 1.5% whereas those that lost more than 1.5% of BW significantly decreased FM by 12%. Both FM and FFM did not significantly change, correspondingly, in those that gain or lost >1.5% BW.
Our findings of a greater contribution of FM to BW loss observed in athletes that lost >1.5% BW are similar to those observed by Garthe et al. (6) who compared the effect of 5% to 6% BW loss with a slow or fast rate on changes in body composition assessed by DXA over 6 months. The changes in BW resulted from FM changes with no significant FFM alteration in the fast rate whereas in the slow rate an increased FFM was even observed from pre to post intervention (6). In our study the season ranged between 5 to 10 month (20 to 40 wk), according to sports, and mean FFM increased by 2.3 kg with FM gains of 0.4 kg in athletes that gained >1.5% BW. Our findings are in the same line as those observed by Garthe et al. (7) that tested the effect of nutritional guidance in ~10 wk of a weight-gain period in elite athletes. In the ad libitum group a nonsignificant mean FM gain of 0.2 kg and a significant mean FFM increase of 1.2 kg were observed.
Deviation from commonly used rules
When adjusted for age, sex, and sports, those that decreased body mass by more than 1.5% showed a contribution of FM and FFM of 90% and 10%, respectively whereas in those that gained more than 1.5% body mass contributions were of 5% and 95%, respectively. Expectedly, the adjusted energy density per BW change was approximately 8700 ± 2147 kcal·kg−1 and 1450 kcal·kg−1, respectively, for those who lost or gained 1.5% of BW.
The smaller and unadjusted contribution of FFM (~15%) and the large impact of FM in the composition of weight change of athletes that lost >1.5% explain the large density of the energy deficit (8200 kcal·kg−1), a value above the Wishnofsky rule (24). The origin of this rule is based on the assumption of a cumulative energy deficit of 7700 kcal·kg−1 of weight loss, involving a fixed energy partition ratio of 87% of FM and 13% of FFM (24).
The energy density of the weight lost observed in our study is also underestimated by the Quarter FFM Rule, established by Webster et al. (23), which states that FFM contribution to BW lost is 25% with a resulting energy of ~7400 kcal·kg−1, based on a regression model using a female group (14–60 yr) with different levels of fatness.
The theoretical equation developed by Forbes quantifies the fat-free proportion of a weight change as a function of the initial body fat. His model has been used as the basis for the recent approaches in dynamic modeling (9,20). In our sample, Forbes’ Curve (4) did not predict changes in FFM accurately, demonstrating a FFM track in the opposite direction from the observed FFM changes occurred over the season. Therefore, Forbes’ Curve is an invalid model for determining the energy partition ratio in an athletic population.
Forbes postulated that changes in FFM are not constant but the FFM fraction of weight loss or gain is greater when initial adiposity is smaller. We did not observe any association between the initial FM with the ratio of FFM/BW changes, assessed by an independent method, DXA. According to Forbes (4), obese individuals (higher baseline FM) under a negative energy balance intervention tend to spare FFM, i.e., a smaller fraction of FFM per each kg of weight change is lost compared to normal weight individuals. In fact, Forbes (4) observed that under a negative energy balance lean subjects (lower baseline FM) tend to lose weight with a higher contribution from FFM. However, we observed that in our sample, composed of lean subjects with lower baseline values of adiposity, FFM was spared, probably due to the vigorous exercise that athletes were exposed to over the season. Garthe and colleagues (6) conducting a weight loss intervention in a sample of elite athletes, either through a fast or slow weight loss rate, observed that FFM was maintained or increased, respectively. Thomas et al. (19) observed that the Forbes curve is valid for describing changes in FFM during caloric restriction but overestimates the decrease in FFM during aerobic exercise.
A number of limitations to the present study should be addressed. First, the athletes were tested at the beginning and at the main competition stage of the season; hence, it is not known if these two measurements are truly representative of a constant energy partition ratio occurring over the entire season.
Second, some of the changes observed in FM were within the propagation of measurement error (<0.7 kg) which may limit the accuracy of this study in detecting small FM changes. It is clear that the main source of the propagation of measurement error was observed from BV estimates, extending the findings of Friedl et al. (5). However, as underlined by Heymsfield et al. (10), the 4C-model improves the accuracy of body fat estimates over that provided by two- and three-component models by accounting for individual variability in TBW and Mo estimates. Still, the important concerns surrounding propagated measurement errors may be reduced by the rapid technological advances that have yet to find their path into applied reference methods (10).
Lastly, as only five sports were included in this study, generalizability of these findings to other sports is limited.
We observed that approximately 95% of the weight gain (>1.5% BW) and nearly 10% of the weight lost (lost >1.5%BW) was comprised by FFM, regardless of sex, age, and sport type. These findings mean that dynamic changes in weight composition of highly trained athletes, specifically if weight is lost, do not follow theoretical driven rules and predictions derived from nonathletic lean or overweight/obese populations.
The authors are grateful to the athletes and coaches for their time and effort.
This work was supported by the Portuguese Foundation for Science and Technology (Grant: PTDC/DES/098963/2008). The current work was also supported by National funding from the Portuguese Foundation for Science and Technology within the R&D units 472 (UID/DTP/00447/2013).
The results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation.
The authors declare that there are no conflicts of interest.
The results of the present study do not constitute endorsement by ACSM.
1. Ackland TR, Lohman TG, Sundgot-Borgen J, et al. Current status of body composition
assessment in sport: review and position statement on behalf of the ad hoc research working group on body composition
health and performance, under the auspices of the I.O.C. Medical Commission. Sports Med
2. Bouchard C, Tremblay A, Després JP, et al. The response to exercise with constant energy intake in identical twins. Obes Res
3. Dulloo AG, Jacquet J. The control of partitioning between protein and fat during human starvation: its internal determinants and biological significance. Br J Nutr
4. Forbes GB. Lean body mass-body fat interrelationships in humans. Nutr Rev
5. Friedl KE, DeLuca JP, Marchitelli LJ, Vogel JA. Reliability of body-fat estimations from a four-compartment model
by using density, body water, and bone mineral measurements. Am J Clin Nutr
6. Garthe I, Raastad T, Refsnes PE, Koivisto A, Sundgot-Borgen J. Effect of two different weight-loss rates on body composition
and strength and power-related performance in elite athletes. Int J Sport Nutr Exerc Metab
7. Garthe I, Raastad T, Refsnes PE, Sundgot-Borgen J. Effect of nutritional intervention on body composition
and performance in elite athletes. Eur J Sport Sci
8. Hall KD. Modeling metabolic adaptations and energy regulation in humans. Annu Rev Nutr
9. Hall KD, Sacks G, Chandramohan D, et al. Quantification of the effect of energy imbalance on bodyweight. Lancet
10. Heymsfield SB, Ebbeling CB, Zheng J, et al. Multi-component molecular-level body composition
reference methods: evolving concepts and future directions. Obes Rev
11. Heymsfield SB, Waki M, Kehayias J, et al. Chemical and elemental analysis of humans in vivo using improved body composition
models. Am J Physiol
. 1991;261(2 Pt 1):E190–8.
12. Heymsfield SB, Wang Z, Baumgartner RN, Ross R. Human body composition
: advances in models and methods. Annu Rev Nutr
13. Merril AL, Watt BK. Energy Value of Foods, basis and derivation (Agriculture Handbook no 74). In: Agriculture Handbook No. 74
. Washington, DC: ARS United States Department of Agriculture; 1973. p. 2.
14. Ross R, Dagnone D, Jones PJ, et al. Reduction in obesity and related comorbid conditions after diet-induced weight loss or exercise-induced weight loss in men. A randomized, controlled trial. Ann Intern Med
15. Santos DA, Silva AM, Matias CN, Fields DA, Heymsfield SB, Sardinha LB. Accuracy of DXA in estimating body composition
changes in elite athletes using a four compartment model as the reference method. Nutr Metab (Lond)
16. Silva AM, Fields DA, Quiterio AL, Sardinha LB. Are skinfold-based models accurate and suitable for assessing changes in body composition
in highly trained athletes? J Strength Cond Res
17. Silva AM, Santos DA, Matias CN, Minderico CS, Schoeller DA, Sardinha LB. Total energy expenditure assessment in elite junior basketball players: a validation study using doubly labeled water. J Strength Cond Res
18. Sundgot-Borgen J, Meyer NL, Lohman TG, et al. How to minimise the health risks to athletes who compete in weight-sensitive sports review and position statement on behalf of the Ad Hoc Research Working Group on Body Composition
, Health and Performance, under the auspices of the IOC Medical Commission. Br J Sports Med
19. Thomas DM, Bouchard C, Church T, et al. Why do individuals not lose more weight from an exercise intervention at a defined dose? An energy balance analysis. Obes Rev
20. Thomas DM, Martin CK, Heymsfield S, Redman LM, Schoeller DA, Levine JA. A Simple Model Predicting Individual Weight Change in Humans. J Biol Dyn
21. Toombs RJ, Ducher G, Shepherd JA, De Souza MJ. The impact of recent technological advances on the trueness and precision of DXA to assess body composition
. Obesity (Silver Spring)
22. Wang Z, Pi-Sunyer FX, Kotler DP, et al. Multicomponent methods: evaluation of new and traditional soft tissue mineral models by in vivo neutron activation analysis. Am J Clin Nutr
23. Webster JD, Hesp R, Garrow JS. The composition of excess weight in obese women estimated by body density, total body water and total body potassium. Hum Nutr Clin Nutr
24. Wishnofsky M. Caloric equivalents of gained or lost weight. Am J Clin Nutr
25. World Medical Association. Declaration of Helsinki: ethical principles for medical research involving human subjects. WMJ; 2008;54:122–5.