A balanced diet with an appropriate energy intake supports optimal body function (Thomas et al., 2016) and is, together with regular physical activity, the cornerstone of a healthy lifestyle. However, exercising women and female athletes focusing on leanness, such as endurance athletes, are reported to be at increased risk of restricted eating behavior and relative energy deficiency related to serious health conditions, including eating disorders, premature osteoporosis, and increased cardiovascular risk factors (De Souza et al., 2014; Mountjoy et al., 2014; Nattiv et al., 2007). There is scientific evidence concerning the causality between relative energy deficiency and the metabolic and endocrine perturbations related to suppressed resting metabolic rate (RMR), subclinical and clinical menstrual dysfunction in women, and poor bone health (Loucks et al., 1998; Loucks & Thuma, 2003). Furthermore, a growing body of evidence suggests that energy deficiency results in an altered endocrine profile, loss of bone mass, and suppressed RMR in male athletes (Dolan et al., 2012; Hagmar et al., 2013; Koehler et al., 2016; Wilson et al., 2015). Nonetheless, recent position papers and reviews call for more knowledge regarding energy deficiency and associated health and performance variables among male athletes (Mountjoy et al., 2014; Tenforde et al., 2016).
RMR represents the energy cost of basic physiological functions, including immunity, reproductive function, growth, and thermoregulation (Fuqua & Rogol, 2013), which all appear to be affected by relative energy deficiency (Mountjoy et al., 2014). When energy intake is inadequate, energy allocation is prioritized to physiological processes essential for immediate survival (Wade & Jones, 2004). Therefore, body weight and body composition may remain within the normal range despite insufficient energy intake (Goldsmith et al., 2010; Redman et al., 2009; Redman & Loucks, 2005). In female athletes, an RMRratio < 0.90 is widely accepted as a surrogate marker for relative energy deficiency (De Souza et al., 2008; Gibbs et al., 2013; Melin et al., 2015).
Traditionally, energy status is evaluated in blocks of 24-hr as either energy balance (EB = energy intake − total energy expenditure) or energy availability (EA = energy intake − exercise energy expenditure [EEE] relative to fat-free mass [FFM]). However, these 24-hr views of human thermodynamics have been criticized for failing to account for the endocrine responses that act on real-time changes in energy intake and expenditure (Benardot, 2013). Within-day EB (WDEB), where energy intake and energy expenditure are assessed in 1-hr intervals may, therefore, be more appropriate (Benardot, 2013; Deutz et al., 2000). Indeed, it has been suggested that failure to find associations between field determinations of low EA and objective measures of energy conservation may be explained by a failure to account for within-day energy deficiency (WDED) as a possible contributor to the metabolic and endocrine alterations associated with relative energy deficiency (Mountjoy et al., 2014). Published studies investigating WDED have thus so far assessed only female athletes, where WDED has been associated with menstrual dysfunction, lower estradiol and RMRratio and higher cortisol levels (Fahrenholtz et al., 2018), and an unfavorable body composition (Deutz et al., 2000).
Therefore, the aim of this study was to estimate and compare WDED, where EB is assessed in 1-hr intervals, in male endurance athletes with suppressed and normal RMR, and to investigate whether these comparisons deviate from the traditional 24-hr assessments. Finally, it was of interest to explore whether WDED is associated with endocrine markers of energy deficiency in this male athletic group.
Methods
A total of 46 male cyclists, triathletes, and long-distance runners were recruited to the study through local clubs and social media in two phases (Figure 1). All subjects were categorized as trained or well trained (Jeukendrup et al., 2000), and at performance levels 3–4 (De Pauw et al., 2013). Inclusion criteria were male, 18–50 years old, absence of disease or injury, maximal oxygen uptake (
—Flowchart showing the recruitment process, dropouts, and exclusion of subjects.
Citation: International Journal of Sport Nutrition and Exercise Metabolism 28, 4; 10.1123/ijsnem.2017-0337
Measurement Methods
Performance and health were assessed during three nonconsecutive days, followed by four consecutive days (three weekdays and one weekend day) of recording food consumption, nonexercise activity thermogenesis (NEAT), and training in the subjects’ normal environment. The test protocol was standardized for each athlete.
On the first day, determination of
Anthropometry
Height measurement was completed without shoes to the nearest 0.1 cm using a centimeter scale affixed to the wall (Seca Optima, Seca, Birmingham, UK), and body weight was measured in light clothing to the nearest 0.01 kg (InBody 720; Biospace, Seoul, Korea). Body mass index was calculated as measured weight in kilograms (kg) divided by height squared in meter (kg/m). Body composition was measured using dual-energy X-ray absorptiometry (Lunar Prodigy, EnCore v. 15; GE Medical Systems, Madison, WI). All measurements were completed in a fasted state between 06:00 and 09:00 a.m.
Maximal Oxygen Uptake
VO2max was predicted by asking the subjects to perform an incremental test until exhaustion: cyclists and triathletes on a stationary bike (Excalibur Sport; Lode B.V., Groningen, The Netherlands) and runners on a treadmill (Katana Sport; Lode B.V.). Cyclists started with 1 min of cycling at a power output corresponding to 3 W/kg, and increased by 25 W/min until voluntary exhaustion or failure to maintain a cadence ≥70 rpm. Runners started at 12 km/hr on a constant incline of 3°. Speed was increased by 1 km·hr−1·min−1 until exhaustion.
Resting Metabolic Rate and Resting Heart Rate
For RMR assessment, subjects arrived at the lab in a fasted state by motorized transport between 06:00 and 09:00 a.m. Subjects were instructed to minimize movement after awakening, and rested lying down for 15 min before the measurements began. For a detailed description of measurement of RMR, see Table 1. The lowest obtained HR during the RMR measurement was registered using a Polar V800 HR monitor (Polar Elektro Oy, Kempele, Finland).
Overview of Methods Used to Calculate WDEB and RMRratio
| Component | Summary of method | Comments/references |
|---|---|---|
| Components of within-day energy balance | ||
| EI | Prospective weighed food and beverage record for four consecutive days in the subjects’ normal environment Digital kitchen scale: OBH Nordica 9843 Kitchen Scale Color, Taastrup, Demark Software program: Dietist Net, Kost och Näringsdata, Bromma, Sweden | In-depth oral and written instructions were given to the subjects |
| DIT | Defined as 10% of EI and distributed in the hours after each meal or snack by using the equation 175.9·T·e−T/1.3, where T = time and e = the base of the natural logarithm | Reed and Hill (1996) |
| NEAT | Subjects wore a Sensewear accelerometer (BodyMedia, Inc., Pittsburgh, PA) or Actigraph accelerometer (Actigraph GT3X®, Pensacola, FL) the same days as dietary intake recording | All logging was performed from the time subjects woke up in the morning until bedtime Only allowed to take the logging device off during showering, swimming, and training |
| EEE | Subjects recorded all training sessions with an HR monitor (Polar M400/V800) during the same days as they recorded dietary intake as epochs of 5 s during every training session EEE (kcal·kg−1·min−1) = ((5.95 × HRaS) + (0.23·age) + (84·1) − 134)/4,186.8, where HRaS = HR above sleeping HR (beats/min) Sleeping HR was estimated from a resting supine measurement during the RMR measurement (sleep HR = 0.83 × supine HR) | Crouter et al. (2008) Brage et al. (2005) |
| EPOC | Defined as 5% of EEE the first hour postexercise plus 3% of EEE the second hour postexercise | Phelain, Reinke, Harris, and Melby (1997) Fahrenholtz et al. (2018) |
| RMR | The pRMR used to calculate WDEB was calculated using the Cunningham equation | Cunningham (1980) |
| SMR | Defined as 90% of pRMR | Used instead of RMR during sleeping hours |
| WDEB | The hourly energy balance was calculated as EB = energy intake − total energy expenditure; predicted DIT + EEE + EPOC + NEAT + RMR | To control for the problem of potential underestimation of energy requirements, the unadapted (pRMR) instead of (mRMR) was used when calculating total energy expenditure The starting point for the calculation of WDEB was at midnight on the first day of food recording and was calculated as follows: the mean EI of the last daily meal/snack minus mean total energy expenditure in the time interval following the mean meal/snack consumption WDEB was calculated continuously for the 4 days of registration |
| WDED variables | Total hours with energy deficit (unadapted EB < 0 kcal) Hours spent in energy deficit exceeding 400 kcal (unadapted EB < −400 kcal) Largest single-hour energy deficit | Benardot (2007) Benardot (2013) |
| Components of RMRratio | ||
| mRMR | Calorimetry using a ventilated canopy hood system (Oxycon Pro; Jaeger GmbH, Hoechberg, Germany) Calibrated before each test according to standards Oxygen consumption and carbon dioxide production were assessed over a 30-min period A 5-min steady state period defined as a CV of less than 10% to assess RMR was identified mRMR was assessed using the Weir equation | Compher et al. (2006) Weir (1990) |
| pRMR | pRMR = 500 + 22 FFM (kg) | Cunningham (1980) |
| RMRratio | RMRratio = mRMR/pRMR Suppressed RMR was defined as a RMRratio < 0.90 and normal RMR as a RMRratio > 0.90 | De Souza et al. (2008) Melin et al. (2015) The Cunningham equation was chosen, because this equation has been found to be the best predictive equation for RMR in endurance athletes (Thompson & Manore, 1996) |
Note. EI = energy intake; RMRratio = resting metabolic rate ratio; pRMR = predicted RMR; mRMR = measured RMR; WDEB = within-day energy balance; SMR = sleeping metabolic rate; DIT = diet-induced thermogenesis; NEAT = nonexercise activity thermogenesis; EEE = exercise energy expenditure; EPOC = excess postexercise oxygen consumption; RMR = resting metabolic rate; SMR = sleeping metabolic rate; CV = coefficient of variation.
Blood Sampling
Fasted blood samples were drawn from a cephalic vein between 07:00 and 09:00 a.m. by a qualified biotechnician. One 10 ml BD Vacutainer CAT (BD, Plymouth, UK) was filled and centrifuged after at least 30 min and within 60 min. Two 1.8 ml Cryotube Vials (Termo Fischer Science, Roskilde, Denmark) were filled with serum and frozen to −75 °C. Blood samples were analyzed for glucose, cortisol, testosterone, and triiodothyronine (T3) at Sørlandets Hospital in Kristiansand and Aker Hormonlab in Oslo, Norway. Reference values based on the Norwegian lab’s standards were used: glucose (4–6 mmol/L); cortisol (138–690 mmol/L); testosterone (18–40 years, 7.2–24 nmol/L; >41 years, 4.6–24 nmol/L); and T3 (1.2–2.7 nmol/L).
Energy Status
EA was calculated by subtracting EEE from the subjects’ daily energy intake, relative to FFM (Nattiv et al., 2007). In order not to underestimate EA, EEE represented only the energy attributable to training, and RMR was subtracted from EEE before being used in the EA calculation.
An overview of the components for the WDEB calculation is presented in Table 1, and an example of WDEB calculation is provided in Table 2, illustrating 18 hr in EB < 0 kcal, 6 hr in EB < −400 kcal, and a largest single-hour deficit of −1,070 kcal.
Example of WDEB Calculation of 1 Day for One Subject
| Kcal in | Kcal out | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Time | EI | DIT | EEE | EPOC | NEAT | SEE | REE | TEE hr to hr | EB |
| 00:00–01:00 | 0 | 0 | 0 | 0 | 0 | 66 | 0 | 66 | 40 |
| 01:00–02:00 | 0 | 0 | 0 | 0 | 0 | 66 | 0 | 1,323 | −26 |
| 02:00–03:00 | 0 | 0 | 0 | 0 | 0 | 66 | 0 | 199 | −92 |
| 03:00–04:00 | 0 | 0 | 0 | 0 | 0 | 66 | 0 | 266 | −158 |
| 04:00–05:00 | 0 | 0 | 0 | 0 | 0 | 66 | 0 | 332 | −224 |
| 05:00–06:00 | 0 | 0 | 0 | 0 | 0 | 66 | 0 | 399 | −290 |
| 06:00–07:00 | 0 | 0 | 0 | 0 | 0 | 66 | 0 | 465 | −356 |
| 07:00–08:00 | 0 | 0 | 0 | 0 | 24 | 0 | 74 | 563 | −454 |
| 08:00–09:00 | 242 | 7 | 0 | 0 | 169 | 0 | 74 | 812 | −462 |
| 09:00–10:00 | 0 | 7 | 0 | 0 | 94 | 0 | 74 | 987 | −637 |
| 10:00–11:00 | 654 | 24 | 0 | 0 | 91 | 0 | 74 | 1,176 | −172 |
| 11:00–12:00 | 13 | 22 | 0 | 0 | 113 | 0 | 74 | 1,384 | −369 |
| 12:00–13:00 | 792 | 38 | 0 | 0 | 163 | 0 | 74 | 1,660 | 148 |
| 13:00–14:00 | 0 | 31 | 0 | 0 | 201 | 0 | 74 | 1,966 | −158 |
| 14:00–15:00 | 575 | 37 | 0 | 0 | 101 | 0 | 74 | 2,178 | 205 |
| 15:00–16:00 | 0 | 28 | 0 | 0 | 162 | 0 | 74 | 2,442 | −59 |
| 16:00–17:00 | 0 | 17 | 721 | 0 | 0 | 0 | 74 | 3,253 | −871 |
| 17:00–18:00 | 278 | 18 | 0 | 36 | 99 | 0 | 74 | 3,481 | −820 |
| 18:00–19:00 | 0 | 12 | 0 | 22 | 142 | 0 | 74 | 3,730 | −1,070 |
| 19:00–20:00 | 1,570 | 55 | 0 | 0 | 88 | 0 | 74 | 3,946 | 283 |
| 20:00–21:00 | 0 | 47 | 0 | 0 | 73 | 0 | 74 | 4,140 | 89 |
| 21:00–22:00 | 259 | 40 | 0 | 0 | 79 | 0 | 74 | 4,333 | 155 |
| 22:00–23:00 | 0 | 27 | 0 | 0 | 79 | 0 | 74 | 4,513 | −25 |
| 23:00–00:00 | 0 | 16 | 0 | 0 | 75 | 0 | 74 | 4,678 | −190 |
| 24-hr Total | 4,383 | 426 | 721 | 58 | 1,753 | 462 | 1,258 | 4,678 | −295 |
Note. WDEB = within-day energy balance; EI = energy intake; DIT = diet-induced thermogenesis; EEE = exercise energy expenditure; EPOC = excess postexercise oxygen consumption; NEAT = nonexercise activity thermogenesis; SEE = sleeping energy expenditure; REE = resting energy expenditure; TEE = total energy expenditure; EB = energy balance.
Statistics
Statistical calculations were performed using RStudio version 0.99.879 (Boston, MA) with a two-tailed significance level of <0.05. All datasets were tested for normality and homogeneity of variance before statistical hypothesis tests were performed. Normally, distributed data were summarized as means and SDs, and nonnormally distributed data as median and interquartile range (25th and 75th percentiles). Differences between subjects with suppressed RMR (RMRratio < 0.90) versus normal RMR (RMRratio > 0.90) were investigated using the unpaired Student’s t test for normally distributed data and the Wilcoxon rank-sum test for nonparametric data. Pearson’s correlation coefficient and Spearman’s rank correlation coefficient were calculated to investigate associations between WDED variables and continuous outcomes for normally and nonnormally distributed data, respectively.
Results
All told, 65% of the subjects had suppressed RMR. Subjects with suppressed RMR were older compared with subjects with normal RMR, but no differences in anthropometry, exercise capacity, training volume (Table 3), or energy expenditure data (Table 4) between the groups were found.
Description of Subjects Characterized by RMRratio
| All (n = 31) | Normal RMR (n = 11) | Suppressed RMR (n = 20) | p valuea | |
|---|---|---|---|---|
| Age (years) | 34.7 ± 8.1 | 30.8 ± 7.2 | 36.9 ± 7.9 | .045 |
| Height (cm) | 179.5 ± 5.3 | 180.4 ± 5.2 | 179.1 ± 5.4 | .516 |
| Body weight (kg) | 72.0 ± 6.1 | 73.7 ± 6.2 | 71.1 ± 6.0 | .267 |
| BMI (kg/m2) | 22.3 ± 1.8 | 22.7 ± 2.1 | 22.1 ± 1.7 | .501 |
| Body fat (kg) | 8.4 (4.5–11.2) | 11.0 (6.0–12.7) | 8.2 (4.2–10.6) | .302 |
| Body fat (%) | 11.7 ± 5.7 | 12.8 ± 6.1 | 11.1 ± 5.5 | .427 |
| Fat-free mass (kg) | 63.4 ± 5.1 | 64.0 ± 4.7 | 63.1 ± 5.4 | .634 |
| Exercise (hr/week) | 8.7 ± 3.2 | 9.2 ± 3.3 | 8.4 ± 3.2 | .515 |
| VO2peak (ml·kg−1·min−1) | 66.4 ± 6.2 | 66.7 ± 8.2 | 66.2 ± 5.0 | .807 |
Note. Data are presented as mean ± SD for normally distributed data and as median and interquartile range (25th–75th percentiles) for nonnormally distributed data. BMI = body mass index; VO2peak = maximal oxygen uptake; RMRratio = resting metabolic rate ratio.
aDifference between subjects with normal (RMRratio > 0.9) versus suppressed RMR (RMRratio < 0.9).
Energy Expenditure and Within-Day Energy Deficiency Characterized by RMRratio
| All (n = 31) | Normal RMR (n = 11) | Suppressed RMR (n = 20) | p valuea | |
|---|---|---|---|---|
| Exercise EE (kcal/day) | 678 ± 250 | 662 ± 283 | 675 ± 238 | .942 |
| DIT (kcal/day) | 228 ± 52 | 243 ± 66 | 220 ± 42 | .250 |
| EPOC (kcal/day) | 54 ± 20 | 55 ± 23 | 54 ± 19 | .920 |
| NEAT (kcal/day) | 819 (482–1,648) | 580 (374–1,094) | 1,548 (557–1,744) | .087 |
| pRMR (kcal/hr) | 79 ± 4 | 79 ± 4 | 79 ± 5 | .741 |
| mRMR (kcal/hr) | 69 ± 8 | 76 ± 8 | 66 ± 5 | <.001 |
| RMRratio | 0.88 ± 0.07 | 0.96 ± 0.05 | 0.83 ± 0.04 | <.001 |
| 24-hr EB (kcal)b | −698 ± 928 | −402 ± 1,056 | −861 ± 832 | .192 |
| 24-hr EB (kcal)c | −914 ± 966 | −463 ± 1,059 | −1,162 ± 837 | .052 |
| 24-hr EA (kcal/kg FFM) | 39 ± 12 | 41 ± 11 | 37 ± 12 | .393 |
| WDEB < 0 kcal (hr/day) | 22.0 (14.1–22.8) | 14.3 (3.9–20.9) | 22.1 (20.4–22.8) | .059 |
| WDEB <−400 kcal (hr/day) | 18.8 (10.5–21.6) | 10.8 (2.5–16.4) | 20.9 (18.8–21.8) | .023 |
| Largest hourly deficit (kcal) | −2,582 ± 2,302 | −1,340 ± 2,439 | −3,265 ± 1,962.9 | .023 |
Note. Data are presented as mean ± SD for normally distributed data and as median and interquartile range (25–75) for nonnormally distributed data. DIT = diet induced thermogenesis; EB = energy balance; EE = energy expenditure; EPOC = excess postexercise oxygen consumption; mRMR = measured resting metabolic rate; pRMR = predicted resting metabolic rate; WDEB = within-day energy balance.
aDifference between subjects with normal (RMRratio > 0.9) versus suppressed RMR (RMRratio < 0.9). bUsing mRMR. cUsing pRMR.
No difference in 24-hr EB or EA between the groups was observed, but subjects with suppressed RMR spent more time in energy deficits exceeding 400 kcal (p = .023) and had larger single-hour energy deficits (p = .023) compared with subjects with normal RMR (Table 4). No difference in protein intake between subjects with normal RMR (1.8 ± 0.4 g·kg−1·day−1) and subjects with suppressed RMR (1.7 ± 0.4 g·kg−1·day−1) was observed.
All subjects had fasting blood glucose, cortisol, testosterone, and T3 within the normal range. There were no associations between WDED and glucose or T3 (Table 5). Larger single-hour energy deficit was associated with higher cortisol (r = .499, p = .004) and a lower testesterone:cortisol ratio (r = .431, p = .015). The more time spent in WDEB < 0 kcal, and the larger the single-hour energy deficit, the lower body fat percentage (r = −.366, p = .043 and r = .359, p = .047, respectively) was found. There were no associations between protein intake and any body composition measures, although there was a tendency toward a lower FFM with lower protein intake (r = −.333, p = .067).
Associations Between Within-Day Energy Deficiency and Markers for Catabolic State
| Hours with WDEB < 0 kcal | Hours with WDEB < −400 kcal | Largest hourly deficit | ||||
|---|---|---|---|---|---|---|
| r | p value | r | p value | r | p value | |
| RMRratio | −.231 | .212 | −.242 | .190 | .335 | .065 |
| Body fat (%) | −.366 | .043 | −.311 | .090 | .359 | .047 |
| Cortisol | .167 | .377 | .294 | .108 | −.499 | .004 |
| Testosterone | −.277 | .132 | −.315 | .085 | .268 | .145 |
| Test:cortisol | −.117 | .532 | −.235 | .204 | .413 | .016 |
| T3 | −.104 | .577 | .032 | .864 | −.058 | .753 |
| Glucose | −.064 | .731 | −.151 | .415 | .247 | .180 |
Note. All subjects (n = 31) were included in the correlation analysis. BP = blood pressure; T3 = triiodothyronine; RMR = resting metabolic rate; Test:cortisol = the ratio between testosterone and cortisol; WDEB = within-day energy balance.
aValues recorded as negative numbers.
Discussion
In this group of well-trained men, 65% had suppressed RMR and, despite similar EA and 24-hr EB compared with subjects with normal RMR, they spent more time in severe energy deficit and had larger single-hour energy deficit, which were associated with higher cortisol levels and a lower testosterone:cortisol ratio.
To account for the endocrine responses, it has been suggested that calculating WDEB is more physiologically relevant compared with the traditional 24-hr assessment (Benardot, 2013). The WDEB method assesses time and magnitude deviations from the predicted EB, where ±400 kcal represent the hypothetical limits for staying in a desirable EB, based on the predicted amount of liver glycogen, although the limits may be smaller or larger, depending on individual factors (Benardot, 2007, 2013; Deutz et al., 2000). Exceeding the threshold of EB below −400 kcal could potentially accelerate catabolic processes and compromise brain glucose availability (Benardot, 2007, 2013). This may be reflected in endocrine alterations, such as higher cortisol levels and lower testosterone:cortisol ratio as observed in our study, which may reduce the ability to recover and increase the risk of overreaching and overtraining, thereby compromising athletic performance (Banfi & Dolci, 2006). WDEB is an accumulating value that does not reset calculations every day at midnight; thus, it is possible that a traditional 24-hr assessment of EB or EA may mask multiday periods with energy deficits. For instance, light training days may have a compensatory effect on the mean 24-hr EB. Such “hidden” periods of energy deficits may, over time, lead to serious health and performance consequences, such as an unfavorable endocrine profile, bone loss, and suppressed RMR (Dolan et al., 2012; Koehler et al., 2016; Wilson et al., 2015).
In an earlier study, the number of hours in EB < −300 kcal was positively associated with body fat percentage in female middle- and long-distance runners (Deutz et al., 2000), presumably related to both an adaptive reduction in RMR and endocrine responses that favor muscle breakdown and fat gain (Benardot, 2007, 2013). Therefore, restrictive eating behavior may have the opposite of the desired effect on athletes’ body composition. This is in contrast with our findings, in which WDED (number of hours in EB < 0 kcal and largest energy deficit) was associated with a lower body fat percentage in male athletes, and, as recently reported, was not associated with body composition in female elite endurance athletes (Fahrenholtz et al., 2018). One explanation for conservation of FFM despite hypocaloric conditions may be attributed to protein intake (Fahrenholtz et al., 2018, Phillips & Van Loon, 2011). This could, however, not explain the findings of this study. The ability to compare our results with those reported by Deutz et al. (2000) is, however, limited due to several methodological differences. For instance, Deutz et al. (2000) used 24-hr recall to assess energy intake and energy expenditure with only one assessment day, in contrast to our four consecutive days of recording food and beverage consumptions as well as objectively measured energy expenditure.
Regarding energy expenditure, some of our athletes had a considerably high NEAT, and although not significantly different, there was a trend toward a higher NEAT in the group with suppressed RMR compared with those with normal RMR. The large NEAT may be due to the fact that some of the athletes were deliberately looking for ways to expend calories to maintain leanness. Another explanation may be the fact that some of the athletes had physically active jobs such as firefighters, carpenters, plumbers, mason workers, and ironworkers. In addition, some athletes self-reported a physically active leisure time such as active play with their children, which to some degree could have increased their NEAT. This information was, however, not registered in the questionnaire, and was obtained only when talking to the athletes. Hence, we can only speculate whether these factors may explain the trend toward a higher NEAT in the group with suppressed RMR. Whether some athletes may not consider their leisure or employment activities as considerably energy-demanding could be an item for future consideration in education programs concerning how to balance energy expenditure with adequate energy intake.
One methodological challenge may be how one distinguishes between “exercise” and “activities that contributed to NEAT.” In our study, detailed information and instructions about the different terms were provided individually to each participant. Exercise was defined as the athletes’ planned exercise bouts, with the aim of improving fitness/performance, while activities that contributed to NEAT were defined as all physical activity besides exercise. Activities such as riding a bike to and from work (if not regarded as an exercise bout by the athletes), walking to the store/in the neighborhood, playing with children, and activity at work (such as working as a plumber or fireman) counted as activities that contributed to NEAT. All athletes were instructed not to use their accelerometer (i.e., to remove it physically from the body) during their planned exercise bouts, and to use their HR monitor during every exercise bout. Accelerometer use is, however, complicated to control (e.g., whether athletes use their accelerometer immediately after exercise), and we recognize that this can have a potential effect on the total NEAT. Detailed information was given to each participant in advance and during data collection to minimize such errors in this study.
When calculating predicted RMR, a prediction error of 10% is expected (Cunningham, 1980), and therefore an expected normal range of RMRratio is 0.9–1.1 (Sterling et al., 2009). A RMRratio < 0.90 has been used as a recognized surrogate marker for energy deficiency in females (De Souza et al., 2008; Gibbs et al., 2013; Melin et al., 2015; Scheid et al., 2009), but more studies are needed to further investigate this relationship in males. Experimental studies indicate that males’ reproductive system may be more resistant to energy deficiency than females’ (Koehler et al., 2016), which may suggest a lower cutoff for RMRratio when assessing male athletes. However, whether males’ sports performance and health consequences other than those related to the reproductive system are similarly sensitive to relative energy deficiency has not yet been investigated. In addition, both the Cunningham and the Harris–Benedict equation have been found to significantly underestimate RMR in heavyweight male national team rower and canoe racers (Carlsohn et al., 2011), suggesting an overestimation of RMRratio in some athletes and an increased risk of false classification of normal RMR with a lower RMRratio cutoff.
Strengths and Limitations
To our knowledge, this is the first study analyzing WDED and associated endocrine markers of energy deficiency in males. Other strengths of this study were inclusion of a relatively high number of male athletes compared with previous research (Carlsohn et al., 2011; Wilson et al., 2015), the use of valid outcome measures, and that all tests followed best-practice protocols for measurements.
The results of this study should be interpreted with consideration of certain methodological limitations. First, the data are based on a cross-sectional study design, limiting assertions of causality. Second, the WDED variables adapted from the literature lead to a high number of correlation analyses, which may increase the risk of a Type I error. Third, a limitation of this study design was that the collection of data related to food consumption, NEAT, and training occurred after the physiological assessment. Hence, we cannot be sure that these behaviors were the cause of the results seen in the study. The reason for assessing dietary intake and energy expenditure after the physiological testing, and not before, was exclusively practical. Because assessment of dietary intake and energy expenditure is methodologically difficult, we needed to give detailed instructions to each participant and ensure that they were all familiar with the measurement equipment and best-practice procedures. In addition, with regard to the participants’ total load of being a part of this project, in combination with their daily life, we chose not to invite them to the lab before the physiological testing. Thus, we decided that the best practical solution was to include the dietary intake and energy expenditure testing after the physiological testing was completed. It should be noted, however, that all athletes were instructed to eat, drink, and exercise “as usual.” The Dietist Net software (Kost och Näringsdata, Bromma, Sweden) was chosen because participants could not see any caloric calculations of their registrations during registration, nor afterward. This may have reduced the risk of under- or over-reporting of food items or portions.
Based on our experiences, we recommend future research to measure dietary intake and energy expenditure immediately before the lab testing to possibly capture a closer correlation between dietary intake and energy expenditure and the physiological variables of interest. Furthermore, there is a need for data that investigate the reasonable period of time over which WDED calculations should be conducted. We also recommend using objective, validated methods to measure both energy intake and energy expenditure, and to standardize when and how the equipment, such as accelerometers or HR monitors, should be used. Finally, to use a registration system that identifies low compliance to the measurement equipment, such as an accelerometer, may be of help to exclude participants not following the test procedures from the analysis. For analysis of energy intake, we recommend using Goldberg’s cutoff (Black, 2000) to reduce the risk of including under-reporters.
In conclusion, we found that male endurance athletes with suppressed RMR, despite similar 24-hr EB and EA, spent more time in energy deficits exceeding 400 kcal and had larger single-hour energy deficits compared with those with normal RMR. WDED was associated with higher cortisol levels and a lower testosterone:cortisol ratio. The results suggest that assessing energy status in intervals of 24 hr may not be sufficient for detecting athletes at risk for health-related consequences caused by energy deficiency. A continuous view on energy status evaluated in smaller time blocks may therefore be more appropriate.
Acknowledgments
The study was funded by the
References
Banfi, G., & Dolci, A. (2006). Free testosterone/cortisol ratio in soccer: Usefulness of a categorization of values. Journal of Sports Medicine & Physical Fitness, 46(4), 611.
Benardot, D. (2007). Timing of energy and fluid intake: New concepts for weight control and hydration. ACSM’s Health & Fitness Journal, 11(4), 13–19. PubMed ID: 29445663 doi:10.1249/01.FIT.0000281226.23643.de
Benardot, D. (2013). Energy thermodynamics revisited: Energy intake strategies for optimizing athlete body composition and performance. Pensar en Movimiento, 11, 1–13. doi:10.15517/pensarmov.v11i2.10841
Black, A.E. (2000). Critical evaluation of energy intake using the Goldberg cut-off for energy intake: Basal metabolic rate. A practical guide to its calculation, use and limitations. International Journal of Obesity, 24(9), 1119–1130. doi:10.1038/sj.ijo.0801376
Brage, S., Brage, N., Franks, P.W., Ekelund, U., & Wareham, N.J. (2005). Reliability and validity of the combined heart rate and movement sensor Actiheart. European Journal of Clinical Nutrition, 59(4), 561–570. doi:10.1038/sj.ejcn.1602118
Carlsohn, A., Scharhag-Rosenberger, F., Cassel, M., & Mayer, F. (2011). Resting metabolic rate in elite rowers and canoeists: Difference between indirect calorimetry and prediction. Annals of Nutrition & Metabolism, 58(3), 239–244. doi:10.1159/000330119
Compher, C., Frankenfield, D., Keim, N., & Roth-Yousey, L. (2006). Best practice methods to apply to measurement of resting metabolic rate in adults: A systematic review. Journal of American Diet Association, 106(6), 881–903. doi:10.1016/j.jada.2006.02.009
Crouter, S.E., Churilla, J.R., & Bassett, D.R. (2008). Accuracy of the Actiheart for the assessment of energy expenditure in adults. European Journal of Clinical Nutrition, 62(6), 704–711. doi:10.1038/sj.ejcn.1602766
Cunningham, J.J. (1980). A reanalysis of the factors influencing basal metabolic rate in normal adults. The American Journal of Clinical Nutrition, 33(11), 2372–2374. PubMed ID: 7435418 doi:10.1093/ajcn/33.11.2372
De Pauw, K., Roelands, B., Cheung, S.S., de Geus, B., Rietjens, G., & Meeusen, R. (2013). Guidelines to classify subject groups in sport-science research. International Journal of Sports Physiology and Performance, 8(2), 111–122. doi:10.1123/ijspp.8.2.111
De Souza, M.J., Nattiv, A., Joy, E., Misra, M., Williams, N.I., Mallinson, R.J., & Matheson, G. (2014). 2014 female athlete triad coalition consensus statement on treatment and return to play of the female athlete triad: 1st International Conference held in San Francisco, California, May 2012 and 2nd International Conference held in Indianapolis, Indiana, May 2013. British Journal of Sports Medicine, 48(4), 289. PubMed ID: 24463911 doi:10.1136/bjsports-2013-093218
De Souza, M.J., West, S.L., Jamal, S.A., Hawker, G.A., Gundberg, C.M., & Williams, N.I. (2008). The presence of both an energy deficiency and estrogen deficiency exacerbate alterations of bone metabolism in exercising women. Bone, 43(1), 140–148. doi:10.1016/j.bone.2008.03.013
Deutz, R.C., Benardot, D., Martin, D.E., & Cody, M.M. (2000). Relationship between energy deficits and body composition in elite female gymnasts and runners. Medicine & Science in Sports & Exercise, 32(3), 659–668. doi:10.1097/00005768-200003000-00017
Dolan, E., McGoldrick, A., Davenport, C., Kelleher, G., Byrne, B., Tormey, W., & Warrington, G.D. (2012). An altered hormonal profile and elevated rate of bone loss are associated with low bone mass in professional horse-racing jockeys. Journal of Bone and Mineral Metabolism, 30(5), 534–542. doi:10.1007/s00774-012-0354-4
Fahrenholtz, I.L., Sjödin, A., Benardot, D., Tornberg, Å.B., Skouby, S., Faber, J., … Melin, A.K. (2018). Within-day energy deficiency and reproductive function in female endurance athletes. Scandinavian Journal of Medicine & Science in Sports, 1–8. doi:10.1111/sms.13030
Fuqua, J.S., & Rogol, A.D. (2013). Neuroendocrine alterations in the exercising human: Implications for energy homeostasis. Metabolism, 62(7), 911–921. doi:10.1016/j.metabol.2013.01.016
Gibbs, J.C., Williams, N.I., & De Souza, M.J. (2013). Prevalence of individual and combined components of the female athlete triad. Medicine & Science in Sports & Exercise, 45(5), 985–996. doi:10.1249/MSS.0b013e31827e1bdc
Goldsmith, R., Joanisse, D.R., Gallagher, D., Pavlovich, K., Shamoon, E., Leibel, R.L., & Rosenbaum, M. (2010). Effects of experimental weight perturbation on skeletal muscle work efficiency, fuel utilization, and biochemistry in human subjects. American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, 298(1), 79–88. doi:10.1152/ajpregu.00053.2009
Hagmar, M., Berglund, B., Brismar, K., & Hirschberg, A.L. (2013). Body composition and endocrine profile of male Olympic athletes striving for leanness. Clinical Journal of Sport Medicine, 23(3), 197–201. doi:10.1097/JSM.0b013e31827a8809
Jeukendrup, A.E., Craig, N.P., & Hawley, J.A. (2000). The bioenergetics of world class cycling. Journal of Science and Medicine in Sport, 3(4), 414–433. doi:10.1016/S1440-2440(00)80008-0
Koehler, K., Hoerner, N.R., Gibbs, J.C., Zinner, C., Braun, H., De Souza, M.J., & Schaenzer, W. (2016). Low energy availability in exercising men is associated with reduced leptin and insulin but not with changes in other metabolic hormones. Journal of Sports Sciences, 34(20), 1921–1929. doi:10.1080/02640414.2016.1142109
Loucks, A.B., & Thuma, J.R. (2003). Luteinizing hormone pulsatility is disrupted at a threshold of energy availability in regularly menstruating women. The Journal of Clinical Endocrinology & Metabolism, 88(1), 297–311. doi:10.1210/jc.2002-020369
Loucks, A.B., Verdun, M., & Heath, E. (1998). Low energy availability, not stress of exercise, alters LH pulsatility in exercising women. Journal of Applied Physiology, 84(1), 37–46. doi:10.1152/jappl.1998.84.1.37
Melin, A., Tornberg, Å.B., Skouby, S., Møller, S., Sundgot-Borgen, J., Faber, J., & Sjödin, A. (2015). Energy availability and the female athlete triad in elite endurance athletes. Scandinavian Journal of Medicine & Science in Sports, 25(5), 610–622. doi:10.1111/sms.12261
Mountjoy, M., Sundgot-Borgen, J., Burke, L., Carter, S., Constantini, N., Lebrun, C., & Ljungqvist, A. (2014). The IOC consensus statement: Beyond the female athlete triad--relative energy deficiency in sport (RED-S). British Journal of Sports Medicine, 48(7), 491–497. doi:10.1136/bjsports-2014-093502
Nattiv, A., Loucks, A.B., Manore, M.M., Sanborn, C.F., Sundgot-Borgen, J., & Warren, M.P. (2007). American College of Sports Medicine position stand. The female athlete triad. Medicine & Science in Sports & Exercise, 39(10), 1867–1882. doi:10.1249/mss.0b013e318149f111
Phelain, J.F., Reinke, E., Harris, M.A., & Melby, C.L. (1997). Postexercise energy expenditure and substrate oxidation in young women resulting from exercise bouts of different intensity. The Journal of the American College of Nutrition, 16, 140–146.
Phillips, S.M., & Van Loon, L.J. (2011). Dietary protein for athletes: From requirements to optimum adaptation. Journal of Sports Sciences, 29, S29–S38. doi:10.1080/02640414.2011.619204
Redman, L.M., Heilbronn, L.K., Martin, C.K., De Jonge, L., Williamson, D.A., Delany, J.P., & Ravussin, E. (2009). Metabolic and behavioral compensations in response to caloric restriction: Implications for the maintenance of weight loss. PloS ONE, 4(2), e4377. doi:10.1371/journal.pone.0004377
Redman, L.M., & Loucks, A.B. (2005). Menstrual disorders in athletes. Sports Medicine, 35(9), 747–755. doi:10.2165/00007256-200535090-00002
Reed, G.W., & Hill, J.O. (1996). Measuring the thermic effect of food. The American Journal of Clinical Nutrition, 63(2), 164–169. doi:10.1093/ajcn/63.2.164
Scheid, J.L., Williams, N.I., West, S.L., VanHeest, J.L., & De Souza, M.J. (2009). Elevated PYY is associated with energy deficiency and indices of subclinical disordered eating in exercising women with hypothalamic amenorrhea. Appetite, 52(1), 184–192. doi:10.1016/j.appet.2008.09.016
Sterling, W.M., Golden, N.H., Jacobsen, M.S., Ornstein, R.M., Hertz, S.M. (2009). Metabolic assessment of menstruating and nonmenstruating normal weight adolescents. International Journal of Eating Disorders, 42, 658–663. doi:10.1002/eat.20604
Tenforde, A.S., Barrack, M.T., Nattiv, A., & Fredericson, M. (2016). Parallels with the female athlete triad in male athletes. Sports Medicine, 46(2), 171–182. doi:10.1007/s40279-015-0411-y
Thomas, D.T., Erdman, K.A., & Burke, L.M. (2016). Position of the academy of nutrition and dietetics, dietitians of Canada, and the American college of sports medicine: Nutrition and athletic performance. Journal of the Academy of Nutrition and Dietetics, 116(3), 501–528. doi:10.1016/j.jand.2015.12.006
Thompson, J., & Manore, M.M. (1996). Predicted and measured resting metabolic rate of male and female endurance athletes. Journal of American Dietetic association, 96(1), 30–34. doi:10.1016/S0002-8223(96)00010-7
Wade, G.N., & Jones, J.E. (2004). Neuroendocrinology of nutritional infertility. American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, 287(6), R1277–R1296. doi:10.1152/ajpregu.00475.2004
Weir, J.B. (1990). New methods for calculating metabolic rate with special reference to protein metabolism. Nutrition, 6(3), 213–221.
Wilson, G., Hill, J., Sale, C., Morton, J.P., & Close, G.L. (2015). Elite male Flat jockeys display lower bone density and lower resting metabolic rate than their female counterparts: Implications for athlete welfare. Applied Physiology, Nutrition, and Metabolism, 40(12), 1318–1320. doi:10.1139/apnm-2015-0354
