Journal Logo


Effect of Increased Dietary Protein on Tolerance to Intensified Training


Author Information
Medicine & Science in Sports & Exercise: April 2011 - Volume 43 - Issue 4 - p 598-607
doi: 10.1249/MSS.0b013e3181f684c9
  • Free


In an attempt to improve endurance performance, athletes often incorporate periods of high training volume, combined with limited recovery time, into carefully planned training regimens (11). A disproportionate balance between training impulse and recovery is likely to result in an accumulation of fatigue, together with perturbations in biological functioning and psychological status (38). Such an imbalance in homeostasis with intensified training ultimately results in an acute impairment in exercise capacity (14) and/or performance (2,12,20).

Endurance athletes interested in enhancing recovery from each intense bout of exercise commonly use nutrition as a tool (15). Carbohydrate (CHO) feeding has been demonstrated to attenuate short-term decrements in endurance performance/capacity after intensified periods of training (2,14). Supplementing a habitual diet for 8 d with CHO before, during, and after (6%, 6%, and 20% CHO solution, respectively) each intense training session (14) or administering a high-CHO diet (8.5 g·kg−1 body mass (BM)·d−1) compared with a moderate-CHO diet (5.4 g·kg−1 BM·d−1) (2) before and during periods of increased training volume resulted in an attenuated impairment in endurance performance. Better maintenance of endurance performance and mood state with CHO feeding during intensified training was associated with attenuated impairments in the stress hormone response to exercise (14) and maintained rates of muscle glycogenolysis (2). In both studies (2,14), protein content of the diet (∼1.9 g·kg−1 BM·d−1) remained constant. CHO has received considerable attention, but despite some evidence, the effect of protein feeding on exercise recovery from an intensified period of training is yet to be investigated.

In a short-term setting, the importance of protein intake during the immediate few hours after endurance exercise has recently received a great deal of attention. The addition of protein to a CHO-rich drink, administered during and after exercise, was reported to improve the endurance capacity of subsequent exercise compared with a CHO-matched but nonisoenergetic drink (27,33). Possible mediators for the improved recovery with additional protein intake include the amelioration of markers of muscle damage, the modulation of postexercise protein balance (21,22), and the initiation of insulin-elicited accelerated rates of glycogen repletion after exercise (39). Moreover, the potential for a high-protein diet to impart a small but beneficial increase in energy substrate availability via gluconeogenic pathways during endurance-type activities cannot be ruled out. Thus, the modulation of multiple physiological responses has been proposed to explain the potentially beneficial role of protein nutrition for optimizing recovery from a strenuous acute bout of endurance exercise.

Given that there may be some benefit of acute ingestion of protein for endurance exercise recovery, it follows that long-term protein ingestion may also be beneficial; however, to date, no study has directly examined the effect of long-term protein feeding on endurance performance in the context of a block of high-intensity training. In a recent field-based study (10), U.S. Marine recruits reported a better maintenance of general health and a reduction in local muscular disruptions when protein was added to nutritional supplements consumed immediately after each training session of a basic 54-d training period compared with those without protein (10). These data provide preliminary evidence that the beneficial role of protein supplementation for exercise recovery observed in short-term-based studies may be extrapolated to the long-term effect of protein feeding on recovery from a period of intensified training. However, a more systematic examination of the response of repeated bouts of exercise to long-term protein feeding is yet to be undertaken.

The purpose of this proof-of-concept study was to examine the effect of additional dietary protein intake during a short-term period of intensified training on subsequent endurance performance. We hypothesized that trained cyclists would experience an improved tolerance to the stress of increased training volume (i.e., fewer perturbations in psychological mood state and markers of endocrine function) when consuming a high-protein diet during and after intensified training, ultimately resulting in the better maintenance of endurance performance.



Eight endurance-trained cyclists (age = 27 ± 8 yr, BM = 73 ± 7 kg; maximal oxygen uptake (V˙O2max) = 64.2 ± 6.5 mL·kg−1·min−1, maximal power output (Wmax) = 372 ± 21 W) were recruited to participate in this study. All cyclists had a training history of at least 5 yr. Before participation, the health status of each participant was assessed using a general health questionnaire. A detailed explanation of practical details, associated risks, and required commitment applicable to the present study preceded the attainment of written informed consent. Participants were reminded of their right to withdraw from the study at anytime without provision of reason. The protocol was approved by the research ethics committee of the School of Sport and Exercise Sciences, University of Birmingham, United Kingdom.

Experimental Design

This study was designed to determine whether increasing dietary protein intake leads to a meaningful attenuation of the short-term decrement in endurance performance, elevated mood disturbance, and perturbations in biological functioning induced by a block of high-intensity training. Each participant completed two trials, both consisting of a 3-wk period of quantified training. Sequentially, training periods were divided into 1 wk of normal training (NOR), 1 wk of intensified training (INT), and 1 wk of recovery training (REC). Endurance performance was assessed by a V˙O2max test and preloaded time trial on days 6 and 7, respectively, of each training week. Identical diets were administered during NOR in both conditions. In a counterbalanced crossover experimental design, participants received either a diet consisting of a high-protein intake (PRO) or a normal protein intake (CON) during INT and REC. Trials were separated by at least a 2-wk washout period in which participants resumed their normal training volume and diet.

Preliminary Testing

Wmax, V˙O2max, submaximal V˙O2 (determined by expired gas collections at each increment of V˙O2max test), and maximal HR (HRmax) were determined within the 7 d before commencing each experimental trial, as previously described (14). All participants completed a familiarization time trial within 7 d before commencement of the study.

Dietary Analysis

In the week preceding the first trial, a 3-d weighed food record was completed. Participants were asked to record accurately all food items they consumed during two weekdays and one weekend day. Diets were analyzed using Internet-based nutrition analysis software ( The habitual daily energy intake of the subjects was 2882 ± 172 kcal, and the habitual diet consisted of the following macronutrient intake: protein = 1.6 ± 0.1 g·kg−1 BM·d−1, CHO = 5.4 ± 0.3 g·kg−1 BM·d−1, and fat = 1.3 ± 0.1 g·kg−1 BM·d−1.

Experimental Trial Periods

Training quantification.

In an attempt to monitor training as accurately as possible and ultimately standardize training impulse between trials, each participant received a cyclists' handbook. On a daily basis, cyclists completed a training diary (time of day training session was started, duration of training (min), average and maximal HR (beats·min−1) for training sessions, location of training, weather conditions, and distance covered (km)) and an assessment of well-being (morning resting HR (RHR; beats·min−1), sleep rating (Likert scale, 1-10), stress rating (Likert scale, 1-10), fatigue rating (Likert scale, 1-10), total sleep (h), and morning BM (kg)). Upon waking, RHR was recorded as the average HR during a 4-min period. On day 6 of each training week, cyclists wore their HR monitors throughout the night (from going to bed at night to getting up in the morning) to record an average sleeping HR (SHR).

Athletes were equipped with a downloadable HR monitor (Polar Vantage NV, Kempele, Finland) for the duration of each trial to monitor individual training sessions. Training zones were calculated in accordance with British Cycling guidelines (1), using lactate threshold (LT) and HR values derived from the preliminary V˙O2max test. Five training zones were defined as follows: zone 1 (Z1) = <70%HRmax, zone 2 (Z2) = 70%-80%HRmax, zone 3 (Z3) = 80%-90%HRmax, zone 4 (Z4) = 90%-95%HRmax, and zone 5 (Z5) = >95%HRmax. During NOR, participants completed their usual volume and type of training. Athletes were required to train on a daily basis during INT, which typically consisted of a combination of high-intensity interval sessions above LT, lasting between 2 and 3 h, and long, continuous rides, usually between 4 and 5 h in duration. Consequently, time spent in Z3-Z5 were increased during INT compared with NOR in both trials. During REC, training volume was reduced to ∼60% relative to NOR in both dietary conditions.

Dietary control.

Diet was controlled for the duration of the study with all food prepared by investigators in the Metabolic Research Kitchen. Five set menus were carefully composed using the Internet-based nutrition analysis software ( Food and drinks were provided as three main meals together with a collection of snacks. Participants were asked to refrain from consuming alcohol and caffeine for the duration of the study.

For each dietary period, participants were fed with the aim to maintain energy balance. The summation of basal metabolic rate (kcal·d−1) (ambulating status), predicted using the Harris Benedict equation and estimated energy expenditure in cycling activity (25), was used to estimate daily energy requirements of cyclists. Total energy intake was adjusted on a daily basis to ensure weight stability. Energy intake was matched between dietary conditions during NOR (PRO/CON = 3711 ± 456 kcal·d−1), INT (PRO = 4409 ± 432 kcal·d−1; CON = 4410 ± 437 kcal·d−1), and REC (PRO = 3096 ± 503 kcal·d−1; CON = 3096 ± 432 kcal·d−1).

The macronutrient composition of the diet during NOR was equivalent to 1.5 g of protein·kg−1 BM·d−1 and 6 g of CHO·kg−1 BM·d−1, with the remainder of energy derived from fat. Hence, for 7 d before a block of high-intensity training, dietary intake and composition were identical between trials. In CON, diet composition remained the same as NOR in both INT and REC, with energy intake modified to match daily changes in energy expenditure. In PRO, during both INT and REC, protein intake was doubled to 3 g·kg−1 BM·d−1. As intended, dietary protein intake was significantly greater in PRO compared with CON during INT (PRO = 210 ± 17 g·d−1; CON = 106 ± 9 g·d−1) and REC (PRO = 210 ± 18 g·d−1; CON = 105 ± 11 g·d−1). The CHO content of the diet was held to 6 g·kg−1 BM·d−1, with the remainder of energy derived from fat. Intake of fat was greater in CON compared with PRO during both INT (PRO = 185 ± 24 g·d−1; CON = 239 ± 30 g·d−1) and REC (PRO = 122 ± 54 g·d−1; CON = 167 ± 52 g·d−1).

In both dietary conditions, CHO-rich fruit smoothies were consumed daily during INT and REC. In PRO, 60 g (3 × 20 g) of casein protein isolate (DSM Food Specialties, Delft, The Netherlands) was added to a fruit smoothie to assist the attainment of increased dietary protein intake. In CON, cream was added to ensure the desired fat intake could be attained. Participants were advised to ingest smoothies on immediate completion of training sessions. All participants confirmed their awareness of which dietary intervention was being administered during each trial.

Performance Tests

Excluding preliminary testing, subjects completed a total of six V˙O2max tests and six time trials to assess maximal power output and endurance performance, respectively. V˙O2max tests and time trials were always performed in the morning (start of exercise between 06:30 and 08:00 a.m.) under similar environmental conditions (∼19°C and ∼55% relative humidity) and in a fasted state (>10 h). The same incremental (workload increased by 35 W at 3-min intervals until volitional exhaustion) maximal cycle ergometer V˙O2max test protocol was performed during experimental trial periods as previously described (14).

Time trial.

To determine endurance performance, a preloaded (120 min of submaximal exercise at 50% Wmax) time trial that lasted approximately 45 min (19) was performed on day 7 of each training week. Previously described general procedures for performance testing were adhered to (8,19). Briefly, on arrival at the laboratory, subjects were fitted with an HR monitor, and a Teflon catheter was inserted into a forearm vein. After a 10-min rest period, a baseline blood sample was drawn. Subjects cycled for 120 min at 50% Wmax with the electromagnetically braked ergometer set in the hyperbolic (cadence-independent) mode so that work rate was independent of pedaling rate. At regular time points (20, 40, 60, 80, and 100 min), blood samples (∼15 mL) were drawn, HR was recorded, and RPE, using the modified Borg scale (4), were obtained. These measurements were repeated immediately after completion of the 120-min submaximal exercise preload (SM). The ergometer was then adjusted to the cadence-dependent (linear) mode. Cyclists commenced the simulated time trial and were required to complete a set amount of work (670 ± 52 kJ) as fast as possible. The total amount of work to be performed was calculated using the following formula (19):

The linear factor was individually adjusted, so that 70% Wmax was obtained when the subject pedaled at their preferred cadence. The only information available to the subjects during the time trial was elapsed work and percentage of work performed (i.e., 0% at the start and 100% on completion). Further blood samples were collected at immediate cessation of time trial and after 1 h of passive recovery.

Blood Analyses

Blood samples were collected into prechilled vacutainers containing K3EDTA or lithium heparin (Becton Dickinson, Franklin Lakes, NJ). Whole blood was immediately placed on ice until centrifugation at 3000 rpm for 10 min at 4°C, within 2 h of collection. Plasma was stored at −80°C until further analysis. Enzymatic analysis for plasma lactate (lactic acid; ABX Diagnostics, Chicksands, UK) and urea concentrations (urea; ABX Diagnostics) were determined in duplicate using a semiautomated analyzer (COBAS MIRA S-plus; ABX Diagnostics). Commercially available sandwich ELISA kits were used to determine plasma cortisol (cortisol; IDS, Tyne and Wear, UK), testosterone (testosterone; IDS), and epinephrine (CAT-COMBO; IDS) concentrations. In each case, plates were read in duplicate on a Labsystems Original Multiskan MS at selected wavelengths (cortisol and testosterone = 450 nm; epinephrine = 405 nm). Reported sensitivities of ELISA were 2.5 ng·mL−1, 0.083 ng·mL−1, and 10 pg·mL−1 for cortisol, testosterone, and epinephrine, respectively. Intra-assay variations were calculated for cortisol (7%), testosterone (8%), and epinephrine (9%). Plasma urea concentrations were determined at all time points during the time trial. Hormone concentrations were measured at baseline, following the submaximal exercise preload and immediately after the time trial. In addition, plasma cortisol was also measured after 1 h of passive recovery. Blood samples collected during each stage of the V˙O2max test were analyzed for plasma lactate concentrations.


Daily analysis of life demands for athletes.

Participants completed the Daily Analysis of Life Demands for Athletes (DALDA) questionnaire (31) on a daily basis to monitor psychological status. The DALDA is divided into parts A and B, which represent the sources and manifestation of stress in the form of signs and symptoms, respectively. For both parts, the number of items marked as "worse than normal" ("a" scores) were tallied and reported.


On day 7 of each training week, before exercise, participants completed a 65-item version of the POMS-65 questionnaire (26). POMS-65 items are divided into six categories: tension, depression, anger, vigor, fatigue, and confusion. Global mood state (GMS) was calculated as the sum of all negative categories minus the score for vigor, plus 100.

Data Presentation and Statistical Analysis

All data are expressed as means ± SD. A two-way (dietary condition and week per day) ANOVA with repeated measures was used as the statistical model to determine differences between dietary conditions, unless otherwise stated. Simple contrasts between diet (two levels: PRO and CON) and week (two levels: NOR and INT, NOR and REC) were selected to generate P values. Plasma urea concentration measured during the time trial and "a" scores on part B of DALDA were expressed as AUC (baseline set as resting values measured during NOR of corresponding dietary condition). Statistical analysis was performed using Statistical Package for Social Sciences (SPSS) 15.0 for Windows (SPSS, Inc., Chicago, IL).

Mechanistic inferences about the population values of statistics were made via magnitude-based precision of estimation, as summarized recently (18). All data sets were tested for normality using Kolmogorov-Smirnov normality tests. If normality was violated, data were transformed using natural log. Precision is presented as ±90% confidence limits (CL) or ×/÷ 90% CL for non-log-transformed and log-transformed data sets, respectively. Effect sizes for log-transformed data sets were calculated by dividing the difference (CON − PRO) of the change (INT − NOR or REC − NOR) in measured parameter by log-transformed SD of CON. Thresholds were held constant at 0.2. For non-log-transformed data sets, the threshold value for small was calculated as 0.2 × SD for CON. In this case, effect size refers to the mean effect calculated as the difference between PRO and CON for the change in value between NOR and INT and NOR and REC.

The likelihood of the outcome being of beneficial, detrimental, or trivial nature was also determined using a published spreadsheet (17); likelihoods were qualified: <1% almost certainly no chance, 1%-5% = very unlikely, 5%-25% = unlikely, 25%-75% = possible, 75%-95% = likely, 95%-99% = very likely, and >99% = almost certain. In the case where most (>50%) of the uncertainty lies between the threshold for a substantially positive and negative effect, the likelihood of the effect being trivial (negligible) is qualified. Effects were described as unclear or inconclusive if confidence intervals overlapped onto both positive and negative values.


Quantification of Training

No clear effect of additional protein intake on weekly training volume was observed during NOR (<1%), INT (<1%), and REC (<1%). Additional protein intake did not clearly affect time spent in each training zone during INT (Z1: 1%, Z2: <1%, Z3: 2%, Z4: 5%, Z5: ∼1.3-fold increase) and REC (Z1: 8%, Z2: 3%, Z3: 9%, Z4: <1%, Z5: <1%).

Time Trial Performance

Figure 1 displays time trial performance after NOR, INT, and REC training in high-protein (PRO) and normal protein (CON) trials. Increased dietary protein intake was unlikely (<1%) harmful and possibly (30%) attenuated (4.3%; CL ×/÷5.4%) the decrement in time trial performance after a block of high-intensity training. Restoration of endurance performance during recovery training possibly (48%) benefited (2.0%; ×/÷4.9%) and was most unlikely (<1%) harmed by additional protein intake.

A, Time trial performance after normal (NOR), intensified (INT), and recovery (REC) training in high-protein (PRO) and normal protein (CON) trials. Values are means ± SD. B, Magnitude of decrement in time trial performance after intensified training (INT) and recovery (REC) compared with normal training in high-protein (PRO) and normal protein (CON) trials. The greater the decrement in performance, the more positive the log-transformed value. Data points refer to back-transformed means, which are fold effects. Dotted lines refer to 90% CI uncertainty bars.

Physiological, Metabolic, and Biochemical Responses

Maximal exercise responses.

No clear effect of PRO was observed for the decrement in Wmax after INT (PRO = 5% ± 4%; CON = 6% ± 4%) and REC (PRO = 1% ± 3%; CON = 1% ± 3%) compared with NOR (PRO = 349 ± 34 W; CON = 357 ± 27 W). Table 1 summarizes the intervention effect of increased dietary protein intake on physiological, metabolic, and biochemical responses to maximal exercise after intensified training. The intervention effect of additional protein intake on modulated V˙O2max values measured after INT (PRO = 4.36 ± 0.11 L·min−1; CON = 4.19 ± 0.73 L·min−1) and REC (PRO = 4.49 ± 0.20 L·min−1; CON = 4.43 ± 0.20 L·min−1) compared with NOR (PRO = 4.24 ± 0.56 L·min−1; CON = 4.49 ± 0.64 L·min−1) were very likely (99.9%) trivial. No clear effects of additional protein intake on maximal HR or maximal lactate concentrations were detected after INT or REC. The decrement in maximal HR (HRmax) (6 ± 1 beats·min−1) and lactate (LACmax) (1.88 ± 0.59 mmol·L−1) responses to exercise after INT (compared with NOR) was not clearly affected by additional protein.

Selected changes in maximal test variables over the course of the study period.

Submaximal exercise and resting responses.

No intervention effect of diet on mean HR (PRO, mean range = 127-128 beats·min−1; CON, mean range = 125-128 beats·min−1) and RPE (PRO, mean range = 12-14; CON, mean range = 11-13) values recorded during the 120-min preload of time trial, RHR (PRO, mean range = 51-52 beats·min−1; CON, mean range = 50-53 beats·min−1) or SHR (PRO, mean range = 50-53 beats·min−1; CON, mean range = 50-51 beats·min−1) was observed. Compared with NOR, increasing dietary protein intake was very likely to substantially increase the plasma urea response to exercise, calculated as AUC, after both INT (PRO = 353; ±164 mmol·L−1; CON = 189; ±76 mmol·L−1) and REC (PRO = 300; ±178 mmol·L−1; CON = 150; ±104 mmol·L−1).

Hormonal responses.

The response of peak stress hormone concentrations to intensified training in PRO and CON are displayed in Figure 2. No clear effect of PRO was detected for the amelioration of lowered peak cortisol concentrations after INT (7.1%; ×/÷6.8%) and REC (7.6%; ×/÷12.9%) compared with NOR. Similarly, the effect of additional protein intake for the amelioration of lowered peak epinephrine concentrations was unclear after INT (13%; ×/÷129%) and REC (16%; ×/÷53%) compared with NOR.

A, Peak cortisol and (B) peak epinephrine concentrations after normal (NOR), intensified (INT), and recovery (REC) training in high-protein (PRO) and normal protein (CON) conditions. Peak cortisol and epinephrine concentrations typically measured immediately following the time trial.

The intervention effect of additional protein intake on modulated peak testosterone concentrations measured after INT (PRO = 7.0 ± 1.7 ng·mL−1; CON = 7.9 ± 2.2 ng·mL−1) and REC (PRO = 6.6 ± 1.4 ng·mL−1; CON = 8.8 ± 2.2 ng·mL−1) compared with NOR (PRO = 7.5 ± 2.2 ng·mL−1; CON = 6.6 ± 1.6 ng·mL−1) were very likely (100%) trivial. No clear effect of increased dietary protein intake on testosterone-cortisol ratio was observed (data not shown).

Mood State and Well-being

Daily analysis of life demands for athletes.

Figure 3 displays the total number of a, i.e., worse than normal, scores on part B of the DALDA questionnaire for both PRO and CON. Notable increases in the number of fatigue-related symptoms including, need for a rest, muscle pains, between-session recovery, general weakness, sleep, and irritability were prevalent after INT compared with NOR in both trials. Increased dietary protein intake led to a very likely (97%) attenuation (17; ±11 AUC of "a" scores part B, DALDA for INT+ REC) of increased symptoms of stress after a block of high-intensity training.

Daily reported number of "worse-than-normal" "a" scores (A) on part B (symptoms of stress) on the DALDA questionnaire during normal (NOR), intensified (INT), and recovery (REC) training and "worse-than-normal" "a" scores (B) on part B (symptoms of stress) on the DALDA questionnaire, expressed as AUC over INT and REC training weeks collectively. Cyclists completed the DALDA questionnaire on waking in a rested state on a daily basis. Values are expressed as means ± SD.

GMS scores: POMS-65 questionnaire.

The subscales of the POMS-65 are presented in Figure 4. Iceberg profiles were flattened, indicating increased stress, after INT in both PRO and CON. Most notable changes after INT were observed in fatigue and vigor. More reported feelings of fatigue were reported after INT compared with NOR in both PRO (NOR = 7 ± 2; INT = 15 ± 1) and CON (NOR = 5 ± 1; INT = 17 ± 1). In contrast, feelings of vigor declined after INT compared with NOR in both PRO (NOR = 16 ± 7; INT = 12 ± 5) and CON (NOR = 16 ± 3; INT = 8 ± 2). However, no clear effect of diet on feelings of fatigue, vigor, depression, anger, and confusion after INT or REC were observed compared with NOR. Likewise, GMS after INT (PRO = 138 ± 30; CON = 144 ± 29) and REC (PRO = 114 ± 36; CON = 104 ± 25) compared with NOR (PRO = 119 ± 36; CON = 110 ± 31) were not clearly affected by additional protein intake.

Iceberg profile (POMS-65) questionnaire after normal (NOR), intensified (INT), and recovery training (REC) in high-protein (PRO; A) and normal protein (CON; B) conditions. Cyclists completed the POMS-65 questionnaire on waking in a rested state on a daily basis. Raw values are expressed as means ± SD.

Psychometric ratings/subjective complaints.

Average weekly fatigue ratings (Likert scale) were higher after INT (5 ± 1) and REC (4 ± <1) compared with NOR (3 ± <1) in CON. A similar pattern was observed for PRO (NOR = 3 ± 1; INT = 6 ± 2; REC = 4 ± 2). No clear effect of diet on fatigue, stress (data not shown), or sleep quality (data not shown) was observed after a block of high-intensity training.


This study was designed as a proof-of-concept investigation to examine the effect of additional protein intake on the change in endurance performance after an intensified period of training. When protein intake was similar to both the habitual intake of recruited cyclists and recently published recommended dietary protein requirements for endurance athletes (1.2-1.4 g·kg−1 BM·d−1) (35), like those in previous studies (2,12,14,20), a substantial decrement in time trial performance and mood state was observed immediately after a block of high-intensity training. Our findings are novel in providing preliminary evidence, suggesting that an increased dietary protein intake of 3 g·kg−1 BM·d−1 attenuates psychological symptoms of stress and may ameliorate decrements in endurance performance after an intensified period of training.

The practical relevance of the attenuated mean decrement in time trial performance after INT when additional dietary protein was administered was examined by an inferential evaluation of our data set using confidence interval-based statistics (17,18). Rather than declaring the value of the statistic based on the probability of a null outcome, use of the analysis of the likelihood of meaningful change has recently been advocated to be a more intuitive and practical statistical approach based directly on uncertainty in the true value of the statistic (4,18). This approach determined that intake of additional protein provides a ∼30% (possible) chance of a beneficial effect and <1% chance of a harmful effect on time trial performance directly after a block of high-intensity training. This inference may be explained by the ∼100-s (∼5%) average attenuation of the decrement in time trial performance designed to last ∼45 min in PRO versus CON after INT compared with NOR. Furthermore, the probability (48%) that attempts to restore endurance performance with 1 wk of recovery training benefited from an increased dietary protein intake existed. Thus, our data provide promising, albeit preliminary, evidence supporting the suggestion that additional protein intake may attenuate decrements in time trial performance after a block of high-intensity training.

It must be acknowledged that our experimental design may have contributed to the somewhat uncertain primary experimental outcome, i.e., the probabilistic term assigned to qualify the effect of additional protein intake on endurance performance only reached possibly beneficial. Practical and logistical considerations precluded tight scientific experimental control. Training load prescribed during the experimental block was not supervised within a laboratory situation; rather, autonomy was given to cyclists to select their own workload within an HR range. Moreover, washout periods between trials were not standardized (range = 2-4 wk). Hence, the possibility exists that these factors combined contributed to an increased variability in physiological responses to dietary intervention and subsequent compromise in reliability of the time trial to assay endurance performance. A coefficient of variation <5% has been calculated for the determination of endurance performance in a glycogen-depleted state using the time trial protocol used in the present study when preceded by a normal training load (7). However, equivalent reliability data do not exist within an intensified training context. In addition, the intensified training model used in the present study has been well established to reduce performance immediately after the intense training period (2,14). Hence, the unexpected observation that one subject improved performance after INT compared with NOR in PRO, exacerbated by a small sample size, may have further reduced the chance of detecting a substantial effect of protein. Thus, it may be argued that our hypothesis that additional protein intake will result in the better maintenance of endurance performance after a high-intensity block of training can neither be comprehensively accepted nor refuted. However, from a purely practical perspective, informing an athlete or coach that a 30% possibly beneficial and <1% most unlikely harmful effect of increasing dietary protein intake on endurance performance during a high-intensity block of training exists may be argued to be attractive.

It should be noted that these results may be applicable only in the context of a relatively low CHO availability. The cyclists in this study consumed a CHO intake equivalent to 6 g·kg−1 BM·d−1 in both trials. This level of CHO intake is at the low end of the recommendation for endurance training and well below that recommended for intense endurance training (6). Thus, increased protein per se may not have been the most salient factor. Previous research has focused on the effect of CHO rather than protein nutrition for reducing the negative effect of intense training. Similar to our results, CHO feeding was shown to alleviate decrements in endurance performance after intensified periods of training by ∼10% (2,14). Depleted muscle glycogen concentrations provide a potential explanation for impaired endurance performance typically experienced after intensified training (34). Because cyclists in the present study consumed a moderate-CHO diet (6 g·kg−1 BM·d−1, i.e., lower than intakes recommended (8-10 g·kg−1 BM·d−1) for intense endurance training (6)), it is likely that muscle glycogen stores were depleted (34), thus contributing to the amelioration of performance decrements in our subjects.

At this suboptimal CHO intake, the increased dietary protein intake may have indirectly increased CHO availability or muscle glycogen stores via gluconeogenic pathways. Support for this notion is exemplified by a study that demonstrated that the rate of gluconeogenesis was elevated by ∼40% with a high-protein diet, consequently accounting for a large proportion of endogenous glucose availability (23). Our urea data may provide indirect evidence that gluconeogenesis was increased during PRO. Urea concentrations were greater with increased protein intake after intensified training, suggesting that more amino acids were deaminated, possibly increasing the availability of carbon skeletons for gluconeogenesis and glycogen synthesis or oxidation to provide energy. In the present study, calculation of glycogen oxidation or the determination of muscle glycogen content could not be made. However, it is possible that the better maintenance of endurance performance after intensified training observed in the high-protein condition may be attributed to a restored rate of CHO oxidation, particularly because CHO intake was suboptimal. Thus, the high-protein intake simply may have partially countered an inadequate CHO availability, rather than having a direct effect per se.

A modulated endocrine response has been proposed as a possible mechanism underpinning the better maintenance of endurance performance when nutrition is manipulated during intensified periods of training (14). The cumulative effect of releasing high concentrations of stress hormones during consecutive bouts of intense exercise has been suggested to desensitize the hypothalamic-pituitary-adrenal axis (16) and ultimately reduce availability of selected hormones (36). This notion is supported by an attenuated peak cortisol response to maximal exercise that we and others (14,37) observed after intensified training. Furthermore, the beneficial effect of CHO nutrition for the attenuation of decrements in performance associated with intensified periods of training has also previously been attributed to a modulated endocrine response (2,14). CHO supplementation before, during, and after training was suggested to attenuate the desensitization of the hypothalamic-pituitary-adrenal axis after intensified training and was thus deemed responsible for restoring the blunted peak cortisol response (14). However, peak stress hormone concentrations measured in the present study were not different between conditions. In addition, peak testosterone concentrations remained within reported "normal" ranges (∼3-10 ng·mL−1) for healthy males (3), irrespective of dietary protein intake. Thus, whereas it is clear that CHO affects the endocrine response to intensified training, the mechanism responsible for the potentially better maintenance of endurance performance with protein feeding does not seem to be related to a modulated stress hormone or testosterone response.

In addition to endocrine parameters, protein nutrition failed to affect physiological and subjective markers of increased training stress. In accordance with previous findings, the number and severity of postexercise physiological markers of stress, such as increased resting and sleeping HR (9), fatigue, and stress ratings, were elevated after consecutive days of excessive training (13). Increasing dietary protein intake during and after intensified training had no effect on these physiological perturbations. Although it cannot be completely ruled out, our data suggest that any improvement in performance with protein feeding is unrelated to physiological or biochemical markers of fatigue or a modulated endocrine response.

The most clear and least speculative explanation responsible for the possibly beneficial effect of additional protein intake on the maintenance of endurance performance during a block of high-intensity training may be related to indirect, centrally driven mechanisms. In agreement with previous findings (2,14,29), psychological symptoms of stress, in particular the number of fatigue-related symptoms of stress (i.e., "need for a rest" and "general weakness"), deteriorated after overload training. CHO feeding has previously been reported (2,14) to attenuate the negative psychological symptoms of stress associated with intensified training. In the present study, this athlete-specific DALDA questionnaire revealed that the number of psychological symptoms of stress reported by cyclists during and after a period of intensified training was attenuated when dietary protein intake was increased. The improved psychological status of athletes during and after a short-term period of intensified training in the high-protein trial potentially contributed to the better maintenance of endurance performance after overload training. Previous research suggests that tyrosine supplementation attenuates decrements in stress-related mood (28); thus, it is possible that an increased availability of amino acids, in particular tyrosine, mediated an improved mood state in the protein condition. The utility of the clinically based POMS questionnaire to monitor the mood state of athletes should be interpreted with caution when applied to an athletic population (29) and therefore may explain why the beneficial effect of protein feeding on mood state was not further substantiated using the POMS. Further research should be conducted to fully elucidate the mechanism(s) responsible for this beneficial effect of protein feeding on psychological status.

Practical implications.

In the context of an intensified period of training, caution should prevail before replicating dietary manipulations administered in the present study. CHO remains the most important fuel to maximize endurance performance during both normal and intensified periods of training. Our experimental design was implemented to maximize the chance of detecting an effect of protein feeding; thus, energy balance during INT was achieved predominantly by increasing protein intake. During strenuous endurance endeavors such as the Tour de France, whereby daily energy expenditures consistently exceed 5000 kcal, protein intakes are regularly reported to be as high as 3 g·kg−1 BM·d−1 (32). However, typically, these athletes concurrently increase CHO intake. In the present study, irrespective of training stress, a standardized moderate-CHO diet (6 g·kg−1 BM·d−1) was consumed, an intake previously demonstrated to be insufficient for maintaining endurance performance during intensified periods of training (2). A recent study (24) demonstrated that, when cyclists maintained normal weekly training volumes, a high-CHO diet rather than an isoenergetic high-protein, moderate-CHO diet resulted in improved endurance performance. Dietary CHO and protein in that study were remarkably similar to that of our subjects during PRO, further supporting the notion that the performance of our subjects may have been influenced by a somewhat inadequate CHO intake. Hence, we acknowledge that the apparent beneficial effect of additional protein intake for the better maintenance of endurance performance was achieved under conditions whereby the CHO intake of cyclists may be deemed suboptimal to support such a short-term period of overload training. Moreover, a recent study (30) demonstrated that 4 h of recovery feeding after two bouts of high-intensity exercise during consecutive days with high-protein feeding imposed over a high-CHO background elicited a delayed (60 h) performance benefit in well-trained male cyclists. Thus, future studies should endeavor to determine whether potentially more realistic protein intakes (e.g., 2 g·kg−1 BM·d−1), which may be more readily practiced by athletes in their natural training setting, combined with recommended CHO intakes (8-10 g·kg−1 BM·d−1) results in better maintenance of endurance performance within the context of an intensified period of training. Furthermore, alternative tests suited to assaying more likely physiological mechanisms of protein dose on performance (i.e., repeat-sprint performance tests) (30) should be considered in future experimental designs.


Our data suggest that increased dietary protein intake may have a beneficial role in exercise recovery by attenuating impairments in endurance performance, which accompany intensified periods of training. It is not clear that protein feeding per se, or the contribution of additional protein toward maintenance of blood glucose or muscle glycogen, was the most important factor. Nevertheless, either way, a likely mediator of this potentially beneficial effect of protein feeding on exercise recovery is perturbations in psychological symptoms of stress. Clearly, further research needs to be done to delineate if protein, per se, is important for recovery from intense exercise and maintenance of performance during periods of intense endurance-based training.

A. K. K. is an employee of DSM Food Specialties. Funding for this work was received from DSM Food Specialties, Delft, The Netherlands.

There are no conflicts of interest for any of the authors.

The results of the present study do not constitute endorsement by American College of Sports Medicine.


1. Achten J, Halson SL, Moseley L, Rayson MP, Casey A, Jeukendrup AE. Higher dietary carbohydrate content during intensified running training results in better maintenance of performance and mood state. J Appl Physiol. 2004;96(4):1331-40.
2. Amory JK, Anawalt BD, Bremner WJ, Matsumoto AM. Daily testosterone and gonadotropin levels are similar in azoospermic and nonazoospermic normal men administered weekly testosterone: implications for male contraceptive development. J Androl. 2001;22(6):1053-60.
3. Batterham AM, Hopkins WG. Making meaningful inferences about magnitudes. Int J Sports Physiol Perform. 2006;1(1):50-7.
4. Borg GA. Perceived exertion: a note on "history" and methods. Med Sci Sports. 1973;5:90-3.
5. Burke L. ACSM and MSSE: nutrition and metabolism perspective. Med Sci Sports Exerc. 2004;36(2):179-85.
6. British Cycling Coach Education Level 2. Club Coach Award Training Course Resources. Leeds (UK): Coachwise Ltd.; 2000. p. 23-4.
7. Currell K, Jentjens RL, Jeukendrup AE. Reliability of a cycling time trial in a glycogen-depleted state. Eur J Appl Physiol. 2006;98(6):583-9.
8. Currell K, Jeukendrup AE. Validity, reliability and sensitivity of measures of sporting performance. Sports Med. 2008;38(4):297-316.
9. Dressendorfer R, Wade C, Schaff J. Increased heart rate in runners: a valid sign of overtraining? Physician Sports Med. 1985;13:77-86.
10. Flakoll PJ, Judy T, Flinn K, Carr C, Flinn S. Postexercise protein supplementation improves health and muscle soreness during basic military training in Marine recruits. J Appl Physiol. 2004;96(3):951-6.
11. Fry RW, Morton AR, Keast D. Periodisation and the prevention of overtraining. Can J Sport Sci. 1992;17(3):241-8.
12. Halson SL, Bridge MW, Meeusen R, et al. Time course of performance changes and fatigue markers during intensified training in trained cyclists. J Appl Physiol. 2002;93(3):947-56.
13. Halson SL, Jeukendrup AE. Does overtraining exist? An analysis of overreaching and overtraining research. Sports Med. 2004;34(14):967-81.
14. Halson SL, Lancaster GI, Achten J, Gleeson M, Jeukendrup AE. Effects of carbohydrate supplementation on performance and carbohydrate oxidation after intensified cycling training. J Appl Physiol. 2004;97(4):1245-53.
15. Hawley JA, Tipton KD, Millard-Stafford ML. Promoting training adaptations through nutritional interventions. J Sports Sci. 2006;24(7):709-21.
16. Heim C, Ehlert U, Hellhammer DH. The potential role of hypocortisolism in the pathophysiology of stress-related bodily disorders. Psychoneuroendocrinology. 2000;25:1-35.
17. Hopkins W. A spreadsheet for deriving a confidence interval, mechanistic inference and clinical inference from a p value. Sportscience. 2007;11:16-9.
18. Hopkins WG, Marshall SW, Batterham AM, Hanin J. Progressive statistics for studies in sports medicine and exercise science. Med Sci Sports Exerc. 2009;41(1):3-13.
19. Jeukendrup A, Saris WH, Brouns F, Kester AD. A new validated endurance performance test. Med Sci Sports Exerc. 1996;28(2):266-70.
20. Jeukendrup AE, Hesselink MK, Snyder AC, Kuipers H, Keizer HA. Physiological changes in male competitive cyclists after two weeks of intensified training. Int J Sports Med. 1992;13(7):534-41.
21. Koopman R, Pannemans DL, Jeukendrup AE, et al. Combined ingestion of protein and carbohydrate improves protein balance during ultra-endurance exercise. Am J Physiol Endocrinol Metab. 2004;287(4):E712-20.
22. Levenhagen DK, Carr C, Carlson MG, Maron DJ, Borel MJ, Flakoll PJ. Postexercise protein intake enhances whole-body and leg protein accretion in humans. Med Sci Sports Exerc. 2002;34(5):828-37.
23. Linn T, Santosa B, Gronemeyer D, et al. Effect of long-term dietary protein intake on glucose metabolism in humans. Diabetologia. 2000;43(10):1257-65.
24. Macdermid PW, Stannard SR. A whey-supplemented, high-protein diet versus a high-carbohydrate diet: effects on endurance cycling performance. Int J Sport Nutr Exerc Metab. 2006;16(1):65-77.
25. McArdle WD, Katch FI, Katch VL. Energy expenditure in household, occupational, recreational, and sports activities. In: Sports and Exercise Nutrition. Baltimore (MD): Lippincott Williams & Wilkins; 1999. p. 589-601.
26. McNair DM. Profile of Mood State Manual. San Diego (CA): Educational and Industrial Testing services; 1971. p. 20.
27. Niles ES, Lachowetz T, Garfi J, et al. Carbohydrate-protein drink improves time to exhaustion after recovery from endurance exercise. Journal of Exercise Physiology. 2001;4(1):45-52.
28. Owasoyo JO, Neri DF, Lamberth JG. Tyrosine and its potential use as a countermeasure to performance decrement in military sustained operations. Aviat Space Environ Med. 1992;63(5):364-9.
29. Rietjens GJ, Kuipers H, Adam JJ, et al. Physiological, biochemical and psychological markers of strenuous training-induced fatigue. Int J Sports Med. 2005;26(1):16-26.
30. Rowlands DS, Rossler K, Thorp RM, et al. Effect of dietary protein content during recovery from high-intensity cycling on subsequent performance and markers of stress, inflammation, and muscle damage in well-trained men. Appl Physiol Nutr Metab. 2008;33(1):39-51.
31. Rushall BS. A tool for measuring stress tolerance in elite athletes. Journal of Applied Sport Psychology. 1990;2:51-66.
32. Saris WH, van Erp-Baart MA, Brouns F, Westerterp KR, ten Hoor F. Study on food intake and energy expenditure during extreme sustained exercise: the Tour de France. Int J Sports Med. 1989;10(1 suppl):S26-31.
33. Saunders MJ, Kane MD, Todd MK. Effects of a carbohydrate-protein beverage on cycling endurance and muscle damage. Med Sci Sports Exerc. 2004;36(7):1233-8.
34. Sherman WM, Doyle JA, Lamb DR, Strauss RH. Dietary carbohydrate, muscle glycogen, and exercise performance during 7 d of training. Am J Clin Nutr. 1993;57(1):27-31.
35. Tarnopolsky M. Protein requirements for endurance athletes. Nutrition. 2004;20(7-8):662-8.
36. Urhausen A, Gabriel H, Kindermann W. Blood hormones as markers of training stress and overtraining. Sports Med. 1995;20(4):251-76.
37. Urhausen A, Gabriel HH, Kindermann W. Impaired pituitary hormonal response to exhaustive exercise in overtrained endurance athletes. Med Sci Sports Exerc. 1998;30(3):407-14.
38. Urhausen A, Kindermann W. Diagnosis of overtraining: what tools do we have? Sports Med. 2002;32(2):95-102.
39. van Loon LJ, Kruijshoop M, Verhagen H, Saris WH, Wagenmakers AJ. Ingestion of protein hydrolysate and amino acid-carbohydrate mixtures increases postexercise plasma insulin responses in men. J Nutr. 2000;130(10):2508-13.


©2011The American College of Sports Medicine