Introduction
An adequate energy balance for athletes helps to ensure the maintenance of body mass (BM) during a training cycle or competitive season, which subsequently may help to optimize training adaptations and recovery. To be in a state of energy balance, total energy intake must match total daily energy expenditure (TDEE). Resting metabolic rate (RMR) provides a substantial contribution to TDEE because several reports have estimated this value to account for approximately 60–70% of TDEE (4,9). Therefore, assessing RMR is an essential component of estimating TDEE, which ultimately will help to identify the daily energy requirements of athletes. Such information can help direct fueling efforts of sport nutrition practitioners. Several factors influence RMR such as energy intake (22), age (16), hormones (16), and nervous system activity (12,18), However, BM, and more specifically body composition, seems to be the primary influential factors regarding an individual's RMR and accounting for the majority of interindividual variances (5,6). Previous results have indicated that fat-free mass (FFM) may account for a large percentage of the variation in RMR compared with other components of BM in both untrained men and women (6,14); however, this relationship becomes less apparent in overweight or obese populations (15). This may in part be explained by a higher contribution of (fat mass [FM]) to RMR because there is evidence that FM does actively consume oxygen, albeit a limited amount (15,20).
It has also been proposed that sex may influence RMR (1,3,18); however, this has yet to be fully examined in an athletic population. Male athletes typically have a higher RMR than female athletes, likely a result of greater body size, and tend to have greater amounts of FFM compared with female athletes, which can also contribute to higher RMR values (6,22). Consequently, because of differences in body composition, it can be hypothesized that differences in absolute RMR may arise. However, it is unknown as to whether these differences would remain when various aspects of body size and composition are controlled. For example, Buchholz et al. (3) found that despite sex differences in absolute RMR, when adjusted for FFM, no differences in RMR were observed in healthy, nonobese adults. Conversely, some studies have found sex differences in RMR to persist even when controlled for FFM (1,18). To date, there is limited research on sex differences in RMR of athletic populations and what factors account for that variation. Thompson and Manore (22) did observe higher absolute RMR values in trained endurance male athletes compared with female endurance athletes; however, when controlled for FFM, the difference in RMR became nonsignificant. The authors also observed that FFM was the single best predictor of individual RMR in male athletes, accounting for 58% of the variance (22). Interestingly, the authors noted that energy intake was the single best predictor of RMR in female athletes accounting for 35% of the variance compared with only 28% of the variation when using FFM (22). Other components of body tissue may also contribute, in varying degrees, to differences in RMR, which may be further influenced by sex. For example, Buchholz et al. (3) reported that FM accounted for a significant amount of the variation in measured RMR in women (28%) but not men (3%). Conversely, Dionne et al. (8) found that FM explained 31% of the variation in 24-hour energy expenditure in men but only 20% in women; meaning that for a given increase in FM, men displayed a larger increase in energy expenditure than women. Clearly, these discrepancies warrant further research in this area to better examining differences in RMR and body composition between sexes. It is possible that because of the unique body composition differences and varying physiques in athletic populations, sex differences in RMR may exist even when adjusted for FFM. This comparison has yet to be examined in athletic populations. Therefore, the purpose of the study was to examine sex differences in RMR in athletic populations when controlling for multiple body composition–related variables. A secondary aim of the study was to identify predictors of RMR.
Methods
Experimental Approach to the Problem
A single day of testing, which included body composition and RMR determination, was used to compare differences in RMR across sex and identify predictors of RMR in an athletic population.
Subjects
Sixty-eight male (mean ± SD: age: 20.1 ± 1.5 years; height: 181.8 ± 5.9 cm; BM: 93.7 ± 16.3 kg; and body fat%: 16.3 ± 8.6%) and 48 female (age: 19.4 ± 1.3 years; height: 166.5 ± 6.0 cm; BM: 63.4 ± 12.7 kg; and body fat%: 21.5 ± 6.3%) National Collegiate Athletic Association (NCAA) Division III athletes participated in this observational study over a single day of testing. Athletes were recruited from the entire athletic department across multiple sports and allowed to participate if they were actively participating in a varsity sport. Participants were instructed to fast >12 hours before testing and refrain from any strenuous exercise or training >48 hours. Exclusion criteria included any athletes currently being treated for or diagnosed with a cardiac, respiratory, circulatory, musculoskeletal, metabolic, immune, autoimmune, psychiatric, hematological, neurological, or endocrinological disorder or disease. Athletes were also instructed to refrain from caffeine >24 hours before testing. This study was approved by the institutional review board at the University of Wisconsin--La Crosse. All athletes were informed of the benefits, risks, and requirements of their participation in the current investigation before signing an institutionally approved informed consent document.
Procedures
Body Composition
Before testing, participants were instructed to refrain from any food for >12 hours with the exception of 1 L of water on waking but no later than within 1 hour of testing. On arrival to the laboratory, participants were assessed for height and BM using a Seca physicians scale for determination of body mass index (BMI). Participants then completed a body composition analysis using a 2 compartment model for body fat percentage (%BF) determination with air displacement plethysmography (BODPOD; Cosmed, Rome, Italy). These variables were used to identify the relationship between different components of BM and RMR. Calibration procedures were completed in accordance with manufacturer guidelines by using an empty chamber and the provided calibrating cylinder of a standard volume (49.55 L) before testing. Participants were instructed to wear tight-fitting clothing, remove all jewelry, and wear a swim cap before entering the testing chamber. Participants were instructed to sit motionless in the chamber using normal breathing patterns. Lung volume was directly assessed to correct for relative body volume. Fat mass and FFM were determined based on the participant's BM and measured body volume obtained using the Brozek equation (2). Test-to-test reliability of performing this body composition assessment in our laboratory with athletic populations has yielded high reliability for BM (r = 0.999), body fat percent (0.994), and FFM (0.998).
Resting Metabolic Rate
Participants then completed a RMR analysis using indirect calorimetry (ParvoMedics True One Metabolic System, Sandy, Utah, USA). This test was a nonexertional test that required the participants to remain in a supine position on an examination table. A clear, hard plastic hood and soft, clear plastic drape were placed over the participant's neck, head, and shoulders to determine resting oxygen uptake and energy expenditure. All participants remained motionless without falling asleep for approximately 20 minutes. Data were recorded after the first 10 minutes of testing during a 5-minute period in which criterion variables (e.g., Vo2 L·min−1) change less than 5%.
Statistical Analyses
All data were analyzed using the Statistical Package for the Social Sciences (SPSS, Version 24.0; SPSS Inc., Chicago, IL, USA). Independent-sample t-tests were used to compare differences in participant demographics, body composition, and RMR-related variables. Zero-order correlations were used to assess the relationships and multicollinearity among variables. Stepwise multiple linear regression analyses were used to determine which predictor variables (BM, FFM, FM, %BF, and BMI) best predicted RMR in men and women and to determine the standard error of estimate (SEE). Alpha was set at p ≤ 0.05 for statistical significance.
Results
As depicted in Table 1, male athletes were reported to have greater amounts of BM, height, FFM, and BMI while also reporting lower %BF levels than female athletes (p < 0.001). Fat mass levels were not different between sexes (p = 0.344).
;)
Table 1.: Descriptive statistics.*
Male athletes had a significantly higher absolute RMR (M: 2,481 ± 209 vs. F: 1,553 ± 193 kcals·d−1; p < 0.001), as seen in Figure 1. When adjusted for BM (M: 25.6 ± 8.3 vs. F: 25.9 ± 2.5 kcals·kg−1 BM per day; p = 0.82) and FFM (M: 31.1 ± 10.6 vs. FM: 33.6 ± 3.8 kcals·kg−1 FFM per day; p = 0.12), this difference became nonsignificant.
Figure 1.: Sex differences in resting metabolic rate (RMR). BM = body mass; FFM = fat-free mass. *Significantly (p > 0.05) greater than females.
Male athletes
The zero-order correlation matrix indicated that BM, FFM, FM, %BF, BMI, and RMR were significantly intercorrelated (r = 0.21–0.92, all p-values < 0.05), as presented in Table 2.
Table 2.: Correlations among physical characteristics and resting metabolic rate (RMR) in male athletes (n = 68).*†
The results of the stepwise multiple linear regression analysis indicated that BM (β = 0.73) significantly contributed to the prediction of RMR, whereas FFM, FM, %BF, and BMI did not.
Female athletes
The zero-order correlation matrix indicated that BM, FFM, FM, %BF, BMI, and RMR were significantly intercorrelated (r = 0.33–0.93, all p values < 0.05), as presented in Table 3.
Table 3.: Correlations among physical characteristics and resting metabolic rate (RMR) in female athletes (n = 48).*†
The results of the stepwise multiple linear regression analysis indicated that BM (β = 0.88) significantly contributed to the prediction of RMR, whereas FFM, FM, %BF, and BMI did not. Therefore, the best-fit equation for the prediction of RMR within the current population is presented in Table 4 for men and women with BM serving as the lone variable.
Table 4.: Equations derived to predict resting metabolic rate (RMR) from body mass (BM) in male (n = 68) and female (n = 48) athletes.*
Discussion
The primary aim of the current study was to examine sex differences in RMR within an athletic population. As expected, male athletes exhibited higher absolute RMR values (M: 2,481 ± 209 vs. F: 1,553 ± 193 kcals·d−1; Figure 1). Sex differences in RMR have been reported within the literature before (1,3); however, the differences are not typically as large as observed in the current study. For example, Buchholz et al. (3) reported higher absolute RMR values in healthy adult men compared with women (M: 1,740 ± 134 vs. F: 1,369 ± 119 kcals·d−1). However, these values only equated to a 39% higher RMR value for men compared with the 60% higher RMR values observed in the current study. Similarly, Arciero et al. (1) observed a 23% higher RMR in healthy adult men compared with women, whereas ten Haaf and Weijs (21) reported a 26% higher RMR in athletic men vs. women. When comparing the participants from these investigations, some meaningful differences must be highlighted. When compared with other studies, the male participants in the current study were substantially larger, with a mean BM of 93.7 ± 16.3 kg, while women were 63.4 ± 12.7 kg, which resulted in a larger discrepancy (∼Δ 30 kg) in BM than observed previously. Previous studies have reported BM values and sex differences of 75.7 ± 7.9 vs. 62.6 ± 6.6 kg (∼Δ 13 kg) (21) and 77.6 ± 11.1 vs. 61.9 ± 8.6 kg (∼Δ 15 kg) (1). To further substantiate this key point, when RMR values were normalized to BM in this study, men and women expended 25.6 ± 8.3 kcals·kg−1 BM per day and 25.9 ± 2.5 kcals·kg−1 BM per day, respectively. These values were similar to the values reported in a cohort of athletic individuals (men: ∼26.5 kcals·kg−1 BM·per day and women: ∼25.4 kcals·kg−1 BM per day) by ten Haff et al. (21). Therefore, it is likely that the greater discrepancy in BM between sexes observed in the current study was an important factor behind the higher absolute RMR values observed and the resulting magnitude of difference in absolute RMR values seen between sexes. In addition, the male athletes in the current study had greater amounts of FFM than previously reported for healthy, active men (∼77 vs. ∼67 kg of FFM) (21), which may have further contributed to the higher RMR values. Previous research has indicated that FFM accounts for a large proportion of the variation in RMR (5,6,9,15,22). As such, higher FFM values have been shown to significantly influence whole body oxygen consumption and account for a large percentage of one's RMR (14). When adjusted for BM and FFM, the differences in RMR became nonsignificant in the current study, which lead us to believe the men have a higher RMR value primarily because of their greater BM and FFM. Because of the contribution of RMR to TDEE, the higher observed RMR values in the current athletic population would certainly lead to higher TDEE levels once the energy expenditure of training, competition, and activities of daily living were considered. This would result in the need for a substantially higher daily caloric intake to achieve a positive, or at the very least, a neutral energy balance.
A secondary aim of the current study was to identify predictors of RMR in athletes. Stepwise linear regression indicated that BM was the strongest predictor of RMR in both men and women, accounting for 53 and 77% of the variation in RMR, respectively. The stepwise addition of height, FFM, FM, and BMI did not further contribute to the predictive ability of RMR for men and women in the current study. This is contradictory to previous findings where FFM has generally been found to be the best predictor of RMR, typically accounting for 60–85% of the interindividual variability (6,13–15,22); however, most previous research involved nonathletic populations or was only validated in athletic populations of varying sport-types, making it difficult to draw comparisons. In favor of the predictive ability of FFM within an athletic population, Thompson and Manore (22) reported that FFM accounted for 58% of the variation in RMR in male endurance athletes, whereas FFM only accounted for approximately 20% of the variation in RMR in the current study for male athletes and 61% of the variation in the female athletes. Similar to Thompson and Manore (22), Cunningham (6) also identified FFM to be a strong predictor of RMR, accounting for nearly 70% of the variability. The predictive favorability of FFM then led to the development of the well-known Cunningham equation, which is popularized for its accuracy in predicting RMR values in athletes (21). However, this equation was initially established using a subset of 239 healthy, nonathletic adults from the data by Harris and Benedict (10) and then validated in athletes. Therein, it may have limited use to certain athletic populations, particularly ones with a larger BM or higher amounts of FFM as seen previously (11). Moreover, the De Lorenzo equation (7) is one of the few RMR prediction equations actually developed from a population of athletes and is derived from the data of 51 male athletes who averaged at least 3 hours of exercise per day. In alignment with the findings of the currently study in that BM functions as a significant predictor of RMR, De Lorenzo et al. (7) also reported that BM was a stronger predictor of RMR compared with FFM in male athletes.
Previous research groups have assessed the predictive ability of several RMR prediction equations and subsequently attempted to develop more population-specific equations of their own. For example, ten Haaf and Weijs (21) collected RMR, height, BM, age, and body composition data on 90 (53M and 37F) adult, Dutch recreational athletes who were exercising, on average, 10.0 ± 5.4 and 7.9 ± 4.0 hours per week for the men and women, respectively. Multiple linear regression analysis was performed on all data to evaluate the performance of several other validated RMR prediction equations and to assess the efficacy of their newly developed equations. Results from this investigation identified the Cunningham equation as having the least error. The authors also noted the newly developed prediction equations, which offered FFM and BM-based versions, could also serve as models to predict RMR because they were similar to the Cunningham equation in terms of accuracy. Specifically, it was found that the Cunningham equation in this group of data accurately predicted the RMR of 84.9% of men and 78.4% of women to within 10% of their measured value. The newly developed equation involving BM, while slightly less, was still able to predict 83% of men and 75.7% of women to within 10% of their measured RMR values. In a similar study, De Lorenzo et al. (7) noted the Mifflin-St. Jeor equation, which is a BM-based equation, accounted for the greatest amount of variance in predicted RMR (77%) and with a SEE of 95 kcal·d−1 compared with other equations, which only accounted for 50–70% of the variance with SEE values of 100–130 kcal·d−1. De Lorenzo et al. (7) also proposed a new BM-based prediction equation that accounted for 78% of the variance in predicted RMR. Previous work from our laboratory indicated that when multiple RMR prediction equations were assessed for their predictive ability in larger male athletes (11), a BM-based equation (Harris-Benedict (10)) was found to have the highest degree of accuracy.
It is difficult to make comparisons across studies because different equations were developed within different populations. As a result, differences in body composition, activity levels, and sex limit the applicability of findings across different sports. These discrepancies and results from previous reports examining the efficacy of different RMR prediction equations are particularly important and salient to our findings from the current study. To start, the study cohort of ten Haaf and Weijs (21) contained athletes from a mixture of sport disciplines in both sexes, while, as mentioned earlier, previous studies have either not included athletes (Cunningham and Harris-Benedict) or have focused entirely on endurance athletes (i.e., De Lorenzo). Even such, the male athletes in this study were notably larger in BM and had higher RMR values than the participants in the ten Haaf's investigation. As previously suggested, this may have contributed to a greater percentage of the variation in RMR because the athletes of the current study were much larger than athletes of previous studies. Specifically, the male athletes in the current study, who were primarily football players (Table 5), had a mean BM of ∼94 kg with an RMR of nearly 2,400 kcal·d−1. Comparatively, athletes in previous studies had a mean BM ranging between 60 and 70 kg, and RMR values of only 1,800 kcal·d−1 (17,19,22), which we propose, may create a scenario where FFM may serve as a better predictor of RMR in smaller and leaner athletes.
Table 5.: Breakdown of sport and training hours per week.
Furthermore, the women in the current study had a mean BM of 63 kg and RMR of 1,500 kcal·d−1 compared with other studies that included female athletes with mean BM ranging from 50 to 55 kg and RMR values of approximately 1,200–1,400 kcal·d−1 (13,22). With increases in BM, there may also be greater amounts of FM, which has previously been found to serve as an independent predictor of RMR (1), contributing to metabolic activities and could also account for the higher RMR values observed in the current study. This phenomenon would also facilitate the scenario where a higher BM may account for an increasingly greater percent of variation in RMR. Therefore, it is plausible that in larger athletes (>85 kg), BM becomes an increasingly stronger predictor of RMR, and BM-based equations may offer higher degrees of accuracy when predicting RMR. Again, this highlights the need for more sport-specific or body-type–specific RMR predictions to be developed, as certain equations may not be applicable across different populations. The results of the regression analysis from the current study provide 2 BM-derived RMR prediction equations, developed in athletes of varying sport type (Table 5). The associated SEE values are slightly higher than those observed with previously developed BM-based RMR prediction equations from athletic populations and when compared with results from a previous study performed by our laboratory group, which assessed the accuracy of several RMR prediction equations (7,11,21). As evidenced, our male and female prediction equations had associated SEE values of 299.3 and 145.0, respectively, whereas, De Lorenzo et al. (7) reported an SEE value of 91 with their BM-based RMR prediction equation, which was developed in male athletes participating in martial arts and water polo. The elevated SEE value is likely attributable to the varying body sizes of athletes used in the current study, which yielded higher a higher variability for BM as seen with SDs twice that of those reported by De Lorenzo et al. (7). Furthermore, the athletes in the current study also had higher RMR values as described previously. Therefore, future research should aim to develop population or sport-specific RMR prediction equations to account for large variations in body composition and levels of training commonly seen with different types of athletes.
Practical Applications
Male athletes are larger in stature with greater amounts of FFM and a lower body fat percentage. Thus, they have higher absolute RMR values that require a greater energy intake to maintain a positive energy balance. Such differences in RMR are primarily the result of body size differences as evidenced by the lack of difference in RMR when adjusted for BM and FFM. Furthermore, we conclude that BM seems to be the strongest predictor of RMR in the current athletic population and have thus provided a novel BM-derived equation to predict RMR. These findings are meaningful to coaches and sport nutrition practitioners because reasonable estimations of energy needs can be made without the completion of expensive, time-consuming methods of body composition assessment. In addition, coaches and practitioners can use this information to improve their own understanding and better educate their athletes on the importance of meeting energy needs. They will better recognize the challenges some athletes may ace in doing this, particularly those athletes with a high BM and who regularly complete high volumes of training.
Acknowledgments
The authors have no conflicts of interest to disclose.
References
1. Arciero PJ, Goran MI, Poehlman ET. Resting metabolic rate is lower in women than in men. J Appl Physiol 75: 2514–2520, 1993.
2. Brozek J, Grande F, Anderson JT, Keys A. Densitometric analysis of body composition: Revision of some quantitative assumptions. Ann N Y Acad Sci 110: 113–140, 1963.
3. Buchholz AC, Rafii M, Pencharz PB. Is resting metabolic rate different between men and women? Br J Nutr 86: 641–646, 2001.
4. Carpenter WH, Poehlman ET, O'Connell M, Goran MI. Influence of body composition and resting metabolic rate on variation in total energy expenditure: A meta-analysis. Am J Clin Nutr 61: 4–10, 1995.
5. Cunningham JJ. A reanalysis of the factors influencing basal metabolic rate in normal adults. Am J Clin Nutr 33: 2372–2374, 1980.
6. Cunningham JJ. Body composition as a determinant of energy expenditure: A synthetic review and a proposed general prediction equation. Am J Clin Nutr 54: 963–969, 1991.
7. De Lorenzo A, Bertini I, Candeloro N, Piccinelli R, Innocente I, Brancati A. A new predictive equation to calculate resting metabolic rate in athletes. J Sports Med Phys Fitness 39: 213–219, 1999.
8. Dionne I, Despres JP, Bouchard C, Tremblay A. Gender difference in the effect of body composition on energy metabolism. Int J Obes Relat Metab Disord 23: 312–319, 1999.
9. Goran MI, Beer WH, Wolfe RR, Poehlman ET, Young VR. Variation in total energy expenditure in young healthy free-living men. Metabolism 42: 487–496, 1993.
10. Harris JA, Benedict FG. A biometric study of human basal metabolism. Proc Natl Acad Sci U S A 4: 370–373, 1918.
11. Jagim AR, Camic CL, Kisiolek J, Luedke J, Erickson J, Jones MT, et al. The accuracy of resting metabolic rate prediction equations in athletes. J Strength Cond Res 32: 1875–1881, 2018.
12. Landsberg L. Insulin resistance, energy balance and sympathetic nervous system activity. Clin Exp Hypertens A 12: 817–830, 1990.
13. Mifflin MD, St Jeor ST, Hill LA, Scott BJ, Daugherty SA, Koh YO. A new predictive equation for resting energy expenditure in healthy individuals. Am J Clin Nutr 51: 241–247, 1990.
14. Nelson KM, Weinsier RL, Long CL, Schutz Y. Prediction of resting energy expenditure from fat-free mass and fat mass. Am J Clin Nutr 56: 848–856, 1992.
15. Nielsen S, Hensrud DD, Romanski S, Levine JA, Burguera B, Jensen MD. Body composition and resting energy expenditure in humans: Role of fat, fat-free mass and extracellular fluid. Int J Obes Relat Metab Disord 24: 1153–1157, 2000.
16. Poehlman ET, Berke EM, Joseph JR, Gardner AW, Katzman-Rooks SM, Goran MI. Influence of aerobic capacity, body composition, and thyroid hormones on the age-related decline in resting metabolic rate. Metabolism 41: 915–921, 1992.
17. Poehlman ET, Melby CL, Badylak SF. Resting metabolic rate and postprandial thermogenesis in highly trained and untrained males. Am J Clin Nutr 47: 793–798, 1988.
18. Poehlman ET, Toth MJ, Ades PA, Calles-Escandon J. Gender differences in resting metabolic rate and noradrenaline kinetics in older individuals. Eur J Clin Invest 27: 23–28, 1997.
19. Schulz LO, Nyomba BL, Alger S, Anderson TE, Ravussin E. Effect of endurance training on sedentary energy expenditure measured in a respiratory chamber. Am J Physiol 260: E257–E261, 1991.
20. Simonsen L, Bulow J, Madsen J. Adipose tissue metabolism in humans determined by vein catheterization and microdialysis techniques. Am J Physiol 266: E357–E365, 1994.
21. ten Haaf T, Weijs PJ. Resting energy expenditure prediction in recreational athletes of 18-35 years: Confirmation of Cunningham equation and an improved weight-based alternative. PLoS One 9: e108460, 2014.
22. Thompson J, Manore MM. Predicted and measured resting metabolic rate of male and female endurance athletes. J Am Diet Assoc 96: 30–34, 1996.
