Dietary practices can influence both acute bone turnover and long-term bone health (38) and feeding influences the diurnal rhythm of bone turnover markers at rest (31). Feeding of a mixed nutrient meal suppresses all markers of bone turnover (4), and feeding of individual nutrients such as glucose, fat, protein, and calcium also suppresses bone resorption at rest (2,3,5,14). However, previous studies have only investigated the response in resting, nonathletic participants who have not performed any prior exercise. It is therefore not known whether there is a similar suppressive effect of nutrient ingestion on bone resorption after exercise in athletic individuals.
Prolonged and intense exercise causes increased bone resorption, as shown by increases in C-terminal telopeptide of type 1 collagen (β-CTX) (19,22,33), although markers of bone formation, such as N-terminal propeptides of type 1 procollagen (P1NP), are less responsive to acute bouts of exercise (13,32,34). Increases in bone resorption, without concomitant increases in bone formation, have been observed for up to 4 d after a bout of exhaustive running (32). Although not definitive, this suggests that prolonged and intense exercise could lead to an uncoupling or imbalance in bone turnover, favoring increased bone resorption, which may have detrimental effects on bone mass and health (15). This uncoupling has been implicated in the formation of stress fracture injuries (30,39), which are debilitating injuries for athletes and on average result in 169 d (with a range of 90 to 270 d) of lost training (23,27). Therefore, maintaining coupled bone turnover and anabolic conditions for bone during and after exercise is important for athletes. Given that endurance athletes train multiple times a day, preventing bone loss and stress fracture injury will help maximize available training time.
Preexercise feeding has been investigated as a potential means for attenuating the bone resorption response to exercise. Scott et al. (34) showed that feeding a mixed nutrient breakfast before exercise had no effect on postexercise β-CTX concentrations compared with fasting, and there were no changes in markers of bone formation. This implies that the mechanical loading experienced during exercise overrides any responses caused by preexercise feeding. Scott et al. (34) also suggested that the stimulatory effect of parathyroid hormone (PTH) on β-CTX may override the effect of preexercise feeding; therefore, other exercise feeding practices, the subsequent PTH response, and related metabolites (calcium and phosphate) require investigation.
Sale et al. (29) showed that carbohydrate (CHO) feeding during exercise attenuated the β-CTX and P1NP responses in the hours after exercise, indicating an acute effect of CHO feeding on bone turnover. However, feeding during intense running is not always well tolerated and is limited by time and practicality. Postexercise feeding provides a practical opportunity to feed multiple nutrients and in the correct amounts, thus allowing athletes to reach other sports nutrition goals, such as aiding muscle glycogen resynthesis, protein synthesis, and maintaining adequate hydration status (16,36), without the restrictions of gastrointestinal discomfort, which commonly limits nutrient ingestion during exercise. Postexercise feeding also allows for investigation of the bone turnover response to acute feeding without the confounding effect of mechanical loading. It is not known whether the acute bone turnover response to postexercise feeding is the same as at rest and whether this varies with different timings of postexercise nutrient ingestion. The aim of this study was to investigate the effect of feeding carbohydrate and protein (CHO + PRO) immediately or 2 h after a prolonged intense running bout, on the bone turnover response in trained endurance runners. Markers associated with exercise and bone were also measured to explore possible mediating and mechanistic factors.
Ten men (mean ± 1 SD, age = 28 ± 6 yr, height = 1.74 ± 0.05 m, body mass [BM] = 69.7 ± 6.3 kg, V˙O2max = 63.0 ± 5.0 mL·kg−1 BM·min−1, weekly running distance = 49.9 ± 12.5 km) completed this study, which was approved by Nottingham Trent University's Ethical Advisory Committee. All participants were trained endurance runners who had been competing and training consistently for a minimum of 2 yr in 10 km, half marathon, marathon, or ultradistance races, without a significant break. Participants had recorded at least one of the following times in the past 2 yr: ≤35 min for 10 km, ≤1:25:00 for half marathon, or ≤3:00:00 for marathon. Participants were recruited from local running and triathlon clubs and local races, via posters, flyers, and posts on club websites. Consent was obtained by the primary researcher. Participants were nonsmokers, had not suffered a fracture in the last 12 months, were free from musculoskeletal injury, and did not suffer from any condition known to affect bone metabolism. Compliance with these inclusion criteria was confirmed in the initial visit to the laboratory where health screening was completed and written informed consent was provided.
This was a randomized (Latin square design), counterbalanced, placebo-controlled, and single-blinded, crossover study. Participants completed a preliminary visit for habituation with trial procedures and measurement of V˙O2max. Participants then completed three 4-d experimental trials, each separated by 1 wk. On days 1 and 2, participants refrained from all exercise and followed a prescribed diet. On day 3, participants performed a bout of treadmill running, at a speed equal to 75% of their previously determined V˙O2max, until volitional exhaustion. Blood samples (20 mL) were collected before exercise, immediately after exercise and every hour after exercise for 4 h. On day 4, participants returned to the laboratory for a fasted follow-up blood sample. The three trials consisted of (i) a placebo (PLA) control trial, where the PLA solution was ingested both immediately and 2 h postexercise; (ii) an immediate feeding (IF) trial, where the CHO + PRO solution was ingested immediately postexercise and the PLA solution 2 h postexercise; and (iii) a delayed feeding (DF) trial, where the PLA solution was ingested immediately postexercise and the CHO + PRO solution 2 h postexercise. In the PLA trial, the CHO + PRO solution was ingested after the final blood sample to ensure that the energy content and the composition of the diet was identical between trials. This meant that a final PLA solution also needed to be ingested in the IF and DF trials to ensure participant blinding to the trial conditions (Fig. 1).
Assessment of V˙O2max
Participants performed an incremental treadmill test to determine lactate threshold, as per Jones and Doust (17), followed by a ramp test to determine V˙O2max. Level running velocities corresponding to 75% V˙O2max (13.0 ± 0.8 km·h−1) were calculated based on the regression of V˙O2 and velocity.
Experimental dietary provision
Participants completed a 3-d food diary for the measurement of habitual energy intake and macronutrient composition. A diet consisting of 55% CHO, 30% fat, and 15% PRO, and isocaloric with habitual diets was designed using dietary analysis software (Nutritics, Dublin, Ireland), for each participant to consume on days 1 and 2 of each trial. Participants provided their own food but were given written and verbal instructions for the preparation of meals, including timings of meals and snacks. Any deviations from prescribed diets were confirmed verbally on day 3 and recorded; there were no significant deviations from prescribed diets.
Experimental trial procedure
Participants were asked to maintain their habitual training and record this throughout the study to help maintain consistency across trials. Participants refrained from all exercise on days 1 and 2. Participants arrived at the laboratory on day 3, after fasting from 2000 h the previous evening and consuming 500 mL of water upon awakening. Shortly after arriving, BM was measured and the first 20 mL blood sample was taken via venipuncture after 10 min of semirecumbent rest.
Participants then ran to volitional exhaustion at 75% V˙O2max, which was preceded by a 5-min warm up and volitional stretching. At exhaustion, a cannula was inserted into a prominent forearm vein, which was kept patent by flushing with saline, a second 20 mL blood sample was taken, with further blood samples taken at 1, 2, 3, and 4 h into recovery. Exact times of exercise commencement, time to exhaustion, and blood samples were recorded and were repeated exactly in each trial within participants to reduce the effect of circadian variation on the results. Because of differences in individual run times to exhaustion between participants, postexercise blood sample timings vary between participants but were controlled within participants. The baseline blood sample was taken at 0840 h and exercise commenced at 0850 h, the blood sample at exhaustion was taken at 10:10 ± 13 min, and blood samples 1–4 h postexercise were taken at 11:10 ± 13 min, 12:10 ± 13 min, 13:10 ± 13 min, and 14:10 ± 13 min.
Depending on the trial, participants were given either the CHO + PRO or PLA solution to consume immediately after exhaustion. Two and 4 h after exhaustion, participants were given further solutions to consume. After the final solution was consumed, participants were provided with food and were free to leave the laboratory. Participants consumed a snack at 1500 h and an evening meal at 1800 h and then remained fasted from 2000 h until the next morning. On day 4, participants arrived in the laboratory after consuming 500 mL of water upon awakening and a final 20 mL blood sample was taken.
Recovery solutions and evening meal composition
The CHO + PRO solution contained 1.5 g·kg−1 BM of CHO (dextrose) and 0.5 g·kg−1 BM of PRO (unflavored whey isolate) that was made up to a 12.5% CHO solution with water. The whey isolate and dextrose mix was tested for banned substances by LGC Supplement Screening (Cambridgeshire, UK; UKAS Testing Laboratory 1187; Certificate of Analysis 91530). Preliminary testing ensured that the PLA solution was taste matched to the CHO + PRO solution using artificial sweetener and flavoring; it consisted of 12 mL·kg−1 BM of water, making this the same volume as the CHO + PRO solution. Participants were blinded to the solutions that they were consuming throughout trials. The total volume of fluid consumed in the three recovery solutions was 2509 ± 227 mL.
On day 3, the overall diet composition was 2000 kcal, 55% CHO, 30% fat, and 15% PRO. The recovery solution contained approximately 500 kcal depending on individual BM; therefore, the snack and evening meal contained approximately 1500 kcal. Deviations from prescribed diets were confirmed verbally on day 4 and recorded; there were no significant deviations from prescribed diets. Participants were allowed to ingest plain water on an ad libitum basis throughout the recovery periods, although none of the participants did this during any of the trials.
Treatment and storage of blood samples
Blood was transferred into precooled tubes and gently inverted five to eight times; 15 mL of blood was transferred into tubes containing 15%, 0.12 mL of K3E EDTA (Becton Dickinson Vacutainer System, Franklin Lakes, NJ), and then centrifuged immediately at 3000 rpm, for 10 min at 5°C, generating plasma. The remaining 5 mL of blood was transferred into standard serum tubes (Becton Dickinson Vacutainer System) and left to clot at room temperature for 60 min before being centrifuged at 2000 rpm for 10 min at 5°C. Plasma and serum was subsequently stored at −80°C until analysis.
β-CTX, P1NP, and PTH were measured by electrochemiluminescence immunoassay on an fully automated COBAS c501 system (Roche Diagnostics, Mannheim, Germany) in blood plasma and were measured in singlicate. The interassay CV for β-CTX was ≤3% between 0.2 and 1.5 μg·L−1, with sensitivity of 0.01 μg·L−1. The interassay CV for P1NP was ≤3% between 20 and 600 μg·L−1, with sensitivity of 8 μg·L−1. The interassay CV for PTH was ≤4% between 1 and 30 pmol·L−1, with sensitivity of 0.8 pmol·L−1. Phosphate (PO4), total calcium (Ca), and albumin were measured in serum, using standard colorimetric assays and spectrophotometric methods, performed on an ABX Pentra 400 (Horiba ABX, Montpellier, France). PO4 was measured using phosphomolybdate, with an inter- and intra-assay CV of ≤3.6% between 0.09 and 7.80 mmol·L−1. Total Ca was measured using ortho-cresolphthalein complexone, with an inter- and intra-assay CV of ≤1.7% between 0.04 and 5.00 mmol·L−1. Albumin was measured using bromocresol green, with an inter- and intra-assay CV of ≤1.9% between 0.02 and 5.99 g·dL−1. Because fluctuations in protein, particularly albumin, may cause total Ca levels to change independently of the ionized calcium (Ca2+) concentration, total Ca concentrations were corrected to give an albumin-adjusted Ca (ACa) value: 0.8 mg·dL−1 was subtracted from the total Ca concentration for every 1.0 g·dL−1 by which the serum albumin concentration was greater than 4 g·dL−1 or 0.8 mg·dL−1 was added to the total Ca concentration for every 1.0 mg·dL−1 by which the serum albumin concentration was less than 4 mg·dL−1; that is, (([albumin] − 4) × −0.8) + [total Ca]. Ca2+, glucose, and lactate were measured in whole blood using a blood gas analyzer (Radiometer ABL90 FLEX, Copenhagen, Denmark). Ca2+ is estimated directly between pH 7.2 and 7.6 with no pH correction applied. The inter- and intra-assay CV values were ≤3% between 0.2 and 9.99 mmol·L−1 for Ca2+, ≤5% between 0 and 60 mmol·L−1 for glucose, and ≤26.7% between 0.1 and 31 mmol·L−1 for lactate.
The study sample size was calculated to detect changes in β-CTX from pre- to postexhaustive exercise, with 85% power at an alpha level of P ≤ 0.05, based on the study by Scott et al. (32). Statistical significance was accepted at an alpha level of P ≤ 0.05. All statistical analyses were performed on raw data. Baseline concentrations were compared using a one-way ANOVA. Parametric assumptions of normality and sphericity were confirmed using Shapiro–Wilk's test and Maulchly's test of sphericity, and where assumptions were violated, a transformation was applied to the data so that the assumptions were satisfied. Normality and homogeneity were achieved after log transformations for ACa and PO4 data. All data were subsequently analyzed using a repeated-measures ANOVA, with trial (PLA vs IF vs DF) and time (of sampling) as within-participant factors. Tukey's HSD post hoc test was used to compare each time point against baseline and to compare trials at each time point. Effect size for multiple comparisons was calculated using partial (
) eta-squared (21). Post hoc comparisons are reported with Cohen's d effect sizes, with d = 0.2 considered as a small effect, d = 0.5 considered as a medium effect, and d = 0.8 considered as a large effect (6). These statistical analyses were performed with Statistica (StatSoft, Tulsa, OK) and SPSS (IBM SPSS Statistics 22, Armonk, NY).
The average time to exhaustion (exercise duration) was 01:15:00 ± 00:13:00. There was a significant decrease in BM from preexercise (69.4 ± 6.1 kg) to postexercise (68.9 ± 5.9 kg) (P = 0.001).
Baseline concentrations of β-CTX, P1NP, PTH, ACa, Ca2+, PO4, and albumin were not significantly different between trials (Table 1).
Habitual Diet and Experimental Dietary Provision
There were no significant differences between the diets prescribed for days 1 and 2 of each trial and the diets that were actually consumed by participants, for overall energy content or macronutrient composition. Participants' habitual diets were not different from the diet provided on day 3 of trials, for overall energy content, CHO content, fat content, and calcium content (P = 0.101–0.523). However, PRO content was significantly higher in the habitual diets compared with the experimental trial diet (P = 0.049) (Table 2).
Bone Turnover Markers
C-terminal telopeptide of type 1 collagen (β-CTX)
There was a significant main effect of trial (P ≤ 0.001,
= 0.581) and time (P ≤ 0.001,
= 0.744) and a significant trial–time interaction (P ≤ 0.001,
= 0.630) for β-CTX. β-CTX concentrations were increased from baseline by the end of exercise in all trials (+8% to +12%). In the PLA trial, β-CTX concentrations remained increased above baseline at 1 h postexercise (+7%), before decreasing thereafter, being significantly lower than baseline concentrations 3 and 4 h postexercise (−31% to −42%, P ≤ 0.001) and 24 h later (−3%). In the IF trial, β-CTX concentrations were significantly lower than baseline at 1 h postexercise and remained below baseline until the end of the trial (−22% to −61%, P ≤ 0.01). In the IF trial, β-CTX concentrations were increased above baseline 24 h later (+8%). In the DF trial, β-CTX concentrations were increased above baseline at 1 h postexercise (+15%), then began to decrease and were significantly lower than baseline concentrations 3 and 4 h postexercise (−44% to −65%, P ≤ 0.001). In the DF trial, β-CTX concentrations were increased above baseline 24 h later (+8%) (Fig. 2A).
At 1 and 2 h postexercise, β-CTX concentrations were significantly lower in the IF trial than the DF (P ≤ 0.001, d = 0.76) and PLA trials (P ≤ 0.001, d = 0.84). At 3 h postexercise, β-CTX concentrations were significantly higher in the PLA trial than the IF (P ≤ 0.001, d = 1.13) and DF trials (P = 0.026, d = 0.54). At 4 h postexercise, β-CTX concentrations were significantly lower in the DF trial than the IF (P = 0.003, d = 0.82) and PLA trials (P ≤ 0.001, d = 1.09) (Fig. 2A). All other time points were not significantly different between trials. The overall β-CTX response was significantly lower in the IF trial than the DF trial (P = 0.019, d = 0.37) and the PLA trial (P ≤ 0.001, d = 0.84).
N-terminal propeptides of type 1 procollagen (P1NP)
There was no main effect of trial for P1NP, but there was for time (P ≤ 0.001,
= 0.621), and there was a significant trial–time interaction (P ≤ 0.001,
= 0.292). P1NP concentrations were significantly increased from baseline by the end of exercise in all trials (+32% to +33%, P ≤ 0.001), and by 1 h, postexercise P1NP had decreased below baseline concentrations in all trials (−3% to −7%). In the PLA trial, P1NP concentrations remained below baseline until the end of the trial (−7% to −9%) but were increased above baseline 24 h later (+4%). In the IF trial, P1NP began to increase and reached concentrations above baseline at 3 and 4 h postexercise (+1% to +3%) and 24 h later (+5%). In the DF trial, P1NP concentrations continued to decrease, and by 3 and 4 h, postexercise were significantly lower than baseline (−10% to −11%, P ≤ 0.05) but were increased above baseline 24 h later (+4%) (Fig. 2B). At 4 h postexercise, P1NP was significantly higher in the IF trial than the DF (P = 0.026, d = 0.20) and PLA trials (P = 0.001, d = 0.25) (Fig. 2B). All other time points were not significantly different between trials.
There was no main effect of trial for PTH, but there was for time (P ≤ 0.001,
= 0.791) and there was a significant trial–time interaction (P ≤ 0.001,
= 0.428). PTH concentrations were significantly increased from baseline by the end of exercise in all trials (+124% to +131%, P ≤ 0.001), but by 1 h, postexercise had decreased significantly below baseline concentrations in all trials (−17% to −37%, P ≤ 0.05). In the PLA trial, PTH concentrations remained below baseline until the end of the trial (−3% to −15%) but were increased above baseline 24 h later (+4%). In the IF trial, PTH then began to increase and reached concentrations above baseline 3 and 4 h postexercise (+2% to +7%) and 24 h later (+1%). In the DF trial, PTH continued to decrease and remained below baseline concentrations for the remainder of the trial (−13% to −27%) and 24 h later (−4%) (Fig. 3A). At 3 h postexercise, PTH was significantly higher in the IF trial than the DF trial (P ≤ 0.001, d = 1.33) (Fig. 3A). All other time points were not significantly different between trials.
Albumin-adjusted calcium (ACa)
There was no main effect of trial for ACa, but there was for time (P = 0.003,
= 0.290), and there was a significant trial–time interaction (P = 0.020,
= 0.191). ACa concentrations were increased from baseline by the end of exercise in all trials (+2% to +3%). In the PLA trial, ACa concentrations remained above baseline until the end of the trial (+2% to +4%) but had decreased below baseline 24 h later (−1%). In the IF trial, ACa remained above baseline (+2% to 3%) until 3 h postexercise when ACa decreased below baseline (−3%), and ACa then increased above baseline 4 h postexercise (+1%) and remained there 24 h later. In the DF trial, ACa remained above baseline until the end of the trial (+2% to +4%) and returned to baseline 24 h later (Fig. 3B). At 3 h postexercise, ACa was significantly lower in the IF trial than the DF (P = 0.008, d = 0.79) and PLA trials (P = 0.001, d = 0.98) (Fig. 3B). All other time points were not significantly different between trials.
Ionized calcium (Ca2+)
There was no main effect of trial for Ca2+, but there was for time (P ≤ 0.001,
= 0.771) and a significant trial–time interaction (P ≤ 0.001,
= 0.321). Ca2+ concentrations were significantly decreased below baseline by the end of exercise in all trials (−5% to −7%, P ≤ 0.001). In the PLA trial, Ca2+ concentrations were still significantly below baseline by 1 h postexercise (−4%, P = 0.002) and remained below baseline until the end of the trial and 24 h later (−3%, P = 0.006). In the IF trial, Ca2+ concentrations had returned to baseline by 1 h postexercise (+1%) and remained at concentrations similar to baseline until the end of the trial and 24 h later (−1%). In the DF trial, Ca2+ concentrations had almost returned to baseline by 1 h postexercise (−1%) and remained at concentrations similar to baseline until the end of the trial and 24 h later (−1%) (Fig. 3C). At 1 h postexercise, Ca2+ concentrations were significantly lower in the PLA trial than the IF trial (P = 0.010, d = 1.41) (Fig. 3C). All other time points were not significantly different between trials.
There was no main effect of trial for PO4, but there was for time (P ≤ 0.001,
= 0.581), and there was a significant trial–time interaction (P = 0.007,
= 0.207). PO4 concentrations were significantly increased above baseline by the end of exercise in all trials (+21% to +26%, P ≤ 0.001). By 1 h postexercise, PO4 concentrations decreased below baseline in all trials (−5% to −13%). In the PLA trial, PO4 concentrations continued to decrease at 2 h postexercise (−8%) then increased and returned to baseline 3 h postexercise. In the PLA trial, PO4 concentrations were increased above baseline at 4 h postexercise (+14%) and 24 h later (+3%). In the IF trial, PO4 concentrations started to increase at 2 h postexercise and increased above baseline 4 h postexercise (+8%). In the IF trial, PO4 concentrations were below baseline 24 h later (−2%). In the DF trial, PO4 concentrations continued to decrease at 2 h postexercise (−8%); concentrations started to increase thereafter but remained below baseline until the end of the trial and 24 h later (−4%) (Fig. 3D). At 1 h postexercise, PO4 concentrations were significantly lower in the IF trial than the DF trial (P = 0.049, d = 1.03) (Fig. 3D). All other time points were not significantly different between trials.
There was no main effect of trial for albumin, but there was for time (P ≤ 0.001,
= 0.372), and there was no trial–time interaction (P = 0.054,
= 0.167). Overall mean albumin concentrations were significantly increased from baseline by the end of exercise (+3% to +4%, P = 0.011). There were no other significant changes in albumin concentrations (Fig. 4).
The main findings of the study are as follows: 1) ingestion of the CHO + PRO solution containing 1.5 g·kg−1 BM of CHO and 0.5 g·kg−1 BM of PRO suppressed β-CTX concentrations after an exhaustive run, with a greater overall suppression when the CHO + PRO solution was ingested immediately; 2) immediate ingestion of the CHO + PRO solution resulted in small increases in P1NP concentrations at 3 and 4 h postexercise; and 3) delayed ingestion of the CHO + PRO solution (2 h postexercise) also resulted in a large suppression of β-CTX concentrations. These findings are novel and have the potential to directly influence an athlete's dietary and/or training practices.
The response in the PLA trial showed that the exhaustive running bout caused an immediate increase in bone turnover at the end of exercise, indicated by increased β-CTX and P1NP concentrations above baseline. This was followed by decreased bone turnover during recovery, indicated by decreased β-CTX and P1NP concentrations below baseline. Ingestion of the CHO + PRO solution immediately postexercise caused a rapid and prolonged (at least 4 h) suppression of β-CTX concentrations below baseline levels (−22% to −61%), whereas ingesting the PLA solution immediately postexercise meant that β-CTX concentrations were increased above baseline by between +7% and +15%. When ingestion of the CHO + PRO was delayed by 2 h, it caused suppression of β-CTX concentrations below baseline (−44 to −65%), which is similar to the suppression caused by immediate ingestion of the CHO + PRO solution and it occurred within the same timeframe, i.e., 1–2 h after ingestion.
This rapid response is important because elite athletes habitually train multiple times a day, meaning that there is often only a few hours in between training sessions and therefore limited time for recovery and food consumption. Although the participants in the present study are not elite athletes, their trained nature means that the results are relevant and may be interpreted and used by elite athletes or practitioners. The results indicate that postexercise nutrient ingestion or exercise commencement can be timed so that the subsequent training session occurs when bone resorption is at its lowest and bone formation at its highest; that is, 3–4 h after the first exercise bout with immediate ingestion of the CHO + PRO solution. This may maximize the anabolic and minimize the catabolic bone response to the subsequent training session; however, further research is needed to investigate whether this intervention does indeed produce a more anabolic environment for bone.
The significant increase in P1NP concentrations (+32% to +33%) and the larger relative increase in P1NP compared with β-CTX concentrations at the end of exercise is interesting, as markers of bone formation are usually less responsive to acute bouts of exercise than markers of bone resorption (13,32,34). Similarly, de Sousa et al. (8) reported a 77% increase in P1NP after a high-intensity interval running session (10 × 800 m). In the present study, P1NP concentrations then decreased to below baseline levels at 1 h postexercise in all trials, but the ingestion of the CHO + PRO solution immediately postexercise caused P1NP to increase above baseline at 3 and 4 h postexercise by between +1% and +3%, whereas ingesting the PLA solution immediately postexercise meant that P1NP remained below baseline concentrations by between −7% and −9%. When the CHO + PRO solution was ingested 2 h postexercise, P1NP concentrations were suppressed further below baseline concentrations (−10% to −11%). It is possible that P1NP could have increased after the last measurement was taken but was missed by the sampling protocol; therefore, it would be useful for future research to examine a longer postexercise period to investigate the longer-term response. The significantly increased P1NP concentration at 4 h postexercise in the IF trial compared with the DF and PLA trials is novel, and taken together, these results advocate the feeding of a CHO + PRO solution immediately postexercise to reduce bone resorption marker concentrations and to increase bone formation marker concentrations in the short-term recovery from intense exercise.
The effects of the CHO + PRO solution did not persist to the morning after exercise, and β-CTX concentrations were increased in the IF and DF trials (+8%) compared with suppressed β-CTX concentrations in the PLA trial (−3%). P1NP was increased 24 h postexercise in all trials (+4% to +5%). This increased bone turnover in the IF and DF trials may reflect the bones adapting to a possible hormonal response that is mediated by feeding. It is unlikely that the bones are adapting to the mechanical loading from the running bout alone, as β-CTX concentrations were not increased 24 h postexercise in the PLA trial. The hormonal mediators of this response are currently unknown; Scott et al. (34) and Sale et al. (29) recently showed that GLP-2, leptin, and ghrelin are unlikely mediators of the effect of CHO or mixed meal feeding on bone turnover. Subsequently, this requires further research including the measurement of other gastro-intestinal hormones.
Although this increased bone turnover response may be positive in subelite populations, elite athletes that train multiple times a day with minimal recovery time and rest days are more likely to suffer from consistently increased bone remodeling, which may have detrimental effects on bone health and enhance the stress fracture risk (25,26,28,30). The trained runners and triathletes in the present study have mean resting bone turnover marker concentrations that are at the upper end of the reference ranges for the nonactive, healthy population (7,11,12). Further, unpublished data from our laboratory show that elite triathletes have mean resting bone turnover marker concentrations that are higher than the trained runners and triathletes. This is supported by Oosthuyse et al. (25) who showed that bone resorption and bone formation markers were significantly elevated each morning after four successive 3 h cycling bouts in well-trained cyclists. Although this is speculative, elite athletes may experience an imbalance between whole-body rates of resorption and formation or defective coupling (26), meaning that neither bone resorption or bone formation is performed adequately and the quality of the bone may be poorer. Alternatively, athletes may experience accelerated remodeling, which can increase bone microdamage accumulation (30), all of which can increase stress fracture risk (1,9,28,30). Indeed it should be noted that in a normal, healthy basic multicellular unit, the suppression of bone resorption may not always be desired if the function of bone resorption is to breakdown and remove damaged bone at areas of microdamage accumulation to allow the area to be repaired and strengthened. Therefore, it is crucial for future research to investigate the long-term effects of postexercise suppression of bone resorption on different athletic and nonathletic populations.
Ingestion of the CHO + PRO solution postexercise is not sufficient to cause a decrease in bone resorption marker concentrations and/or an increase in bone formation marker concentrations 24 h postexercise. However, as elite athletes rarely go 24 h without a training session and often have a second session within 4 h of finishing the first session, the bone turnover response 24 h postexercise is less important than the immediate response as it does not reflect real life athlete practice. The more important time point is therefore 4 h postexercise, as this may be around the same time that the second training session would start. As we have now investigated the effect of postexercise feeding after a single acute bout of exercise, future studies should investigate the effect of postexercise nutrient ingestion on repeated bouts of exercise occurring on the same day.
The responses of Ca2+ and PO4 to exercise are in line with previous research (37), and the responses are only significantly different between trials at 1 h postexercise; Ca2+ concentrations were lower in the PLA trial compared with the IF trial, suggesting that IF augments the recovery of Ca2+ to baseline concentrations, and PO4 is lower in the IF trial compared with the DF trial. Transient peaks in PTH concentrations, as shown in the present study, are shown to be anabolic for bone (10), and Townsend et al. (37) showed that PTH secretion during exercise and recovery is controlled by both Ca2+ and PO4. Therefore, these metabolites are likely to be mediating any anabolic effect of increased PTH concentrations. The fact that PTH and P1NP follow the same response could suggest that PTH is mediating an anabolic response in the IF trial; however, this response needs to be confirmed.
At 3 h postexercise, PTH concentrations were greater in the IF trial compared with the DF trial (+7% vs −27%). This response coincides with significantly lower ACa concentrations at 3 h postexercise in the IF trial compared with the DF and PLA trials (−3% vs +3% to +4%). β-CTX concentrations were at their lowest at 3 h postexercise in the IF trial. Considering that the action of increased PTH secretion is to increase calcium through mobilization of the bone reservoir via activation of bone resorption (and also by increasing renal tubular reabsorption and intestinal calcium absorption) (24,35,40), this suggests that changes in PTH and calcium metabolism are unlikely to mediate the acute suppression in bone resorption seen with postexercise CHO + PRO feeding. However, ACa has been shown to be unsuitable when investigating the rapid response of calcium metabolism to exercise (37), which may also be true when investigating CHO + PRO ingestion around exercise. Although Ca2+ (nonprotein bound calcium) decreased at the end of exercise, because albumin concentrations increased, ACa was normalized and remained fairly unchanged throughout exercise. Changes in albumin could have been effected by the ingestion of dietary protein throughout the recovery period, which has previously been shown to increase circulating albumin concentrations (18,20); however, albumin did not change significantly throughout the recovery period. The increase in albumin at the end of exercise could have been to encourage more calcium to be transported around the body because of the tissues requiring additional Ca2+ to keep up with the demand in energy consumption, although the increase in albumin might also just reflect hemoconcentration as a result of the running bout. Transient hemoconcentration can occur rapidly after the onset of acute exercise, possibly even occurring before any significant losses of fluid through sweating or respiration, and it might be argued that significant hemoconcentration would mean that changes in plasma solutes simply reflect shifts in plasma volume. However, one might argue that the level of a plasma solute, irrespective of plasma volume shifts, is more important because it is this that the body responds to. The data presented herein are uncorrected for plasma volume changes, which could influence the interpretation of the biological data obtained during the recovery period, and this should be considered when interpreting results. It is recommended that future studies take this into consideration and correct bone turnover marker data for plasma volume shifts, where appropriate, perhaps even presenting these data both corrected and uncorrected for plasma volume changes.
In conclusion, after exhaustive running, immediate ingestion of a CHO + PRO recovery solution may be beneficial as it decreases bone resorption marker concentrations and increases bone formation marker concentrations, creating a more positive bone turnover balance. The mechanisms underlying the acute changes in bone turnover remain unknown, but a change in calcium metabolism is unlikely to fully mediate the response.
Professor Craig Sale has received funding from the GlaxoSmithKline Human Performance Laboratory for other studies relating to nutrition and bone health. The authors have no professional relationships with companies or manufacturers who will benefit from the results of the present study. The results presented herein do not constitute endorsement by the American College of Sports Medicine. The results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation.
1. Bennell KL, Malcolm SA, Thomas SA, et al. Risk factors for stress fractures in track and field athletes. A twelve-month prospective study. Am J Sports Med
2. Bjarnason NH, Henriksen EE, Alexandersen P, Christgau S, Henriksen DB, Christiansen C. Mechanism of circadian variation in bone resorption. Bone
3. Blumsohn A, Herrington K, Hannon RA, Shao P, Eyre DR, Eastell R. The effect of calcium supplementation on the circadian rhythm of bone resorption. J Clin Endocrinol Metab
4. Clowes J, Hannon R, Yap T, Hoyle N, Blumsohn A, Eastell R. Effect of feeding on bone turnover markers and its impact on biological variability of measurements. Bone
5. Clowes JA, Allen HC, Prentis DM, Eastell R, Blumsohn A. Octreotide abolishes the acute decrease in bone turnover in response to oral glucose. J Clin Endocrinol Metab
6. Cohen J. Statistical Power Analysis for the Behavioral Sciences
. 2nd ed. Hillsdale (NJ): Lawrence Erlbaum Associates; 1988. p. 567.
7. de Papp AE, Bone HG, Caulfield MP, et al. A cross-sectional study of bone turnover markers in healthy premenopausal women. Bone
8. de Sousa MV, Pereira RM, Fukui R, Caparbo VF, da Silva ME. Carbohydrate beverages attenuate bone resorption markers in elite runners. Metabolism
9. Fredericson M, Jennings F, Beaulieu C, Matheson GO. Stress fractures in athletes. Top Magn Reson Imaging
10. Frolik CA, Black EC, Cain RL, et al. Anabolic and catabolic bone effects of human parathyroid hormone (1-34) are predicted by duration of hormone exposure. Bone
11. Glover SJ, Garnero P, Naylor K, Rogers A, Eastell R. Establishing a reference range for bone turnover markers in young, healthy women. Bone
12. Glover SJ, Gall M, Schoenborn-Kellenberger O, et al. Establishing a reference interval for bone turnover markers in 637 healthy, young, premenopausal women from the United Kingdom, France, Belgium, and the United States. J Bone Miner Res
13. Guillemant J, Accarie C, Peres G, Guillemant S. Acute effects of an oral calcium load on markers of bone metabolism during endurance cycling exercise in male athletes. Calcif Tissue Int
14. Henriksen DB, Alexandersen P, Bjarnason NH, et al. Role of gastrointestinal hormones in postprandial reduction of bone resorption. J Bone Miner Res
15. Ihle R, Loucks AB. Dose–response relationships between energy availability and bone turnover in young exercising women. J Bone Miner Res
16. Jentjens R, Jeukendrup A. Determinants of post-exercise glycogen synthesis during short-term recovery. Sports Med
17. Jones AM, Doust JH. A 1% treadmill grade most accurately reflects the energetic cost of outdoor running. J Sports Sci
18. Kaysen GA, Gambertoglio J, Jimenez I, Jones H, Hutchison FN. Effect of dietary protein intake on albumin homeostasis in nephrotic patients. Kidney Int
19. Kerschan-Schindl K, Thalmann M, Sodeck GH, et al. A 246-km continuous running race causes significant changes in bone metabolism. Bone
20. Kirsch R, Frith L, Black E, Hoffenberg R. Regulation of albumin synthesis and catabolism by alteration of dietary protein. Nature
21. Lakens D. Calculating and reporting effect sizes to facilitate cumulative science: a practical primer for t-tests and ANOVAs. Front Psychol
22. Maïmoun L, Manetta J, Couret I, et al. The intensity level of physical exercise and the bone metabolism response. Int J Sports Med
23. Matheson GO, Clement DB, McKenzie DC, Taunton JE, Lloyd-Smith DR, MacIntyre JG. Stress fractures in athletes. A study of 320 cases. Am J Sports Med
24. McSheehy PM, Chambers TJ. Osteoblast-like cells in the presence of parathyroid hormone release soluble factor that stimulates osteoclastic bone resorption. Endocrinology
25. Oosthuyse T, Badenhorst M, Avidon I. Bone resorption is suppressed immediately after the third and fourth days of multiday cycling but persistently increased following overnight recovery. Appl Physiol Nutr Metab
26. Parfitt AM. The coupling of bone formation to bone resorption: a critical analysis of the concept and of its relevance to the pathogenesis of osteoporosis. Metab Bone Dis Relat Res
27. Ranson CA, Burnett AF, Kerslake RW. Injuries to the lower back in elite fast bowlers: acute stress changes on MRI predict stress fracture. J Bone Joint Surg Br
28. Riggs BL, Melton LJ 3rd, O'Fallon WM. Drug therapy for vertebral fractures in osteoporosis: evidence that decreases in bone turnover and increases in bone mass both determine antifracture efficacy. Bone
29. Sale C, Varley I, Jones TW, et al. Effect of carbohydrate feeding on the bone metabolic response to running. J Appl Physiol (1985)
30. Schaffler MB, Radin EL, Burr DB. Long-term fatigue behavior of compact bone at low strain magnitude and rate. Bone
31. Schlemmer A, Hassager C. Acute fasting diminishes the circadian rhythm of biochemical markers of bone resorption. Eur J Endocrinol
32. Scott JP, Sale C, Greeves JP, Casey A, Dutton J, Fraser WD. The effect of training status on the metabolic response of bone to an acute bout of exhaustive treadmill running. J Clin Endocrinol Metab
33. Scott JP, Sale C, Greeves JP, Casey A, Dutton J, Fraser WD. The role of exercise intensity in the bone metabolic response to an acute bout of weight-bearing exercise. J Appl Physiol (1985)
34. Scott JP, Sale C, Greeves JP, Casey A, Dutton J, Fraser WD. Effect of fasting versus feeding on the bone metabolic response to running. Bone
35. Thorsen K, Kristoffersson A, Hultdin J, Lorentzon R. Effects of moderate endurance exercise on calcium, parathyroid hormone, and markers of bone metabolism in young women. Calcif Tissue Int
36. Tipton KD, Elliott TA, Cree MG, Wolf SE, Sanford AP, Wolfe RR. Ingestion of casein and whey proteins result in muscle anabolism after resistance exercise. Med Sci Sports Exerc
37. Townsend R, Elliott-Sale KJ, Pinto AJ, et al. Parathyroid hormone secretion is controlled by both ionised calcium and phosphate during exercise and recovery in men. J Clin Endocrinol Metab
38. Walsh JS, Henriksen DB. Feeding and bone. Arch Biochem Biophys
39. Warden SJ, Burr DB, Brukner PD. Stress fractures: pathophysiology, epidemiology, and risk factors. Curr Osteoporos Rep
40. Zittermann A, Sabatschus O, Jantzen S, Platen P, Danz A, Stehle P. Evidence for an acute rise of intestinal calcium absorption in response to aerobic exercise. Eur J Nutr