Dietary protein ingestion acts synergistically with resistance exercise to elevate rates of muscle protein synthesis (MPS) (1, 2) and enhance whole-body net protein balance (3), which over time supports training adaptations such as muscle hypertrophy. The importance of dietary protein for the recovery from and adaptation to resistance exercise is reflected in the generally accepted greater daily recommendations in athletic populations relative to their weight-stable, non-training counterparts (4). For example, traditional nitrogen balance methodology has estimated protein requirements to be ~1.33 g·kg⁻¹·d⁻¹ (5), which is greater than the current recommended daily allowance of 0.8 g·kg⁻¹·d⁻¹ (6). However, the safe protein intake (i.e., +2 SD of the estimated average requirement; EAR) by nitrogen balance has also been suggested to be greater in novice weight lifters as compared to trained body builders (i.e., ~1.7 vs. 1.2 g·kg⁻¹·d⁻¹) (7, 8). This may be related in part to an increased requirement to support muscle damage repair and growth that is typically greatest during the initiation of a resistance training program (9, 10). Viewed through this lens, recent estimates of the protein requirements to maximize whole body protein synthesis on a non-training day by the indicator amino acid oxidation technique (IAAO) in experienced body builders, who would likely have limited hypertrophic potential, of 1.7 g·kg⁻¹·d⁻¹ (safe intake at the upper 95% confidence interval of 2.2 g·kg⁻¹·d⁻¹) could be interpreted as curiously high (11). Nevertheless, this EAR is similar to our recent estimates using the IAAO in females (i.e., ~1.5 g·kg⁻¹·d⁻¹) (12) and men [~1.9 g·kg⁻¹·d⁻¹; (13)] on a training day. However, these estimated requirements may be irrelevant considering strength athletes habitually consume well in excess of these recommended intakes (i.e., >2 g·kg⁻¹·d⁻¹) (5), which was indeed the case with the aforementioned IAAO studies that reported mean habitual intakes of ~1.9–2.4 g·kg⁻¹·d⁻¹ (11–13).

The ability to incorporate dietary amino acids into muscle and body proteins is saturable, with meal protein intakes in excess of ~0.3 g·kgBW-1 leading to increased amino acid oxidation (14–16). Overtime, the consumption of protein above the maximal anabolic threshold results in an upregulation of amino acid oxidative capacity as an essential component of the adaptive metabolic demand for protein (17, 18). Studies estimating protein requirements by nitrogen balance typically provide a controlled test diet intake of 5–11d to allow for adaptation to a new intake in light of this adaptive upregulation in amino acid oxidative capacity (19). In contrast, IAAO has been suggested to be valid with only a 2-d pre-trial controlled diet (20), which is a purported strength of the technique permitting a greater number and range of protein intakes to be tested (21). However, the IAAO method relies on the oxidation of a tracer amino acid (the reciprocal of whole-body protein synthesis) (22) and therefore any metabolic adaptations (e.g., upregulation of oxidative enzymes) associated with a high habitual protein intake may impact the oxidation of the indicator or other amino acids, which would translate into an overestimation of protein requirements. This could explain, in part, the relatively high protein recommendations recently reported in resistance trained populations with high habitual protein intakes utilizing IAAO methodology (11–13). Therefore, the purpose of this study was to determine what impact an acute reduction in protein intake has in high habitual protein consumers on ¹³CO₂ excretion (F¹³CO₂, the primary outcome of IAAO trials) and the time course for any subsequent metabolic adaptations. In addition, lean mass growth with training has been reported to be supported by a range of protein intakes (i.e., ~1.6 ± 0.4 g·kg⁻¹·d⁻¹) (23) with moderate levels (i.e., 1.2–1.4 g·kg⁻¹·d⁻¹) sufficient to support anabolism in novice weightlifters at the onset of training (i.e., ~12wk) (24, 25). Thus, a secondary outcome was to determine how reducing dietary protein intake to a moderate 1.2 g·kg⁻¹·d⁻¹ from a high intake of 2.2 g·kg⁻¹·d⁻¹ could support post-exercise anabolism over a 5-d adaptation period. This moderate intake was selected as it represents the IAAO-derived safe intake for non-exercising adults (26), the lower 95% confidence interval from (11) in bodybuilders on a non-training day, is the intake often used for a 2 days adaptation period (12, 27, 28), and is within the range of intakes that have been reported to support lean mass growth with training (23). We hypothesized that, compared to a high protein intake, F¹³CO₃ would be elevated and whole-body net protein balance would be reduced during post-exercise recovery for up to 5 days after reducing protein intake to a moderate intake in resistance trained athletes accustomed to a high habitual protein diet.

Dietary amino acids consumed in excess of their ability to be incorporated into de novo muscle and body protein synthesis are irreversibly oxidized (22), representing the primary contributor to the adaptive metabolic demand for amino acids (39). We demonstrate that reducing dietary protein intake in habitually high protein consumers from a standardized 2.2 g·kg⁻¹·d⁻¹ to a moderate 1.2 g·kg⁻¹·d⁻¹ resulted in a reduction in whole-body net protein balance during recovery from resistance exercise for up to 3 days, although greater durations (i.e., ≥5 days) may be needed for full adaptation. Importantly, this reduced net balance was not related to any accommodation in whole body protein breakdown (i.e., PheRa) but rather was the result of an elevated excretion of our indicator amino acid (i.e., F¹³CO₂), suggesting that metabolic adaptation in high habitual protein consumers requires greater durations than the traditional 2 days adaptation period employed in IAAO studies (11, 12, 20, 30).

The principle of the IAAO is the indicator amino acid (phenylalanine) is partitioned toward oxidation when there is a limitation in a single or multiple amino acids (with the exception of tyrosine) to support its use for protein synthesis. Seminal work in the development of the IAAO method demonstrated that 8 h to 9 days of adaptation to various intakes of single amino acids (i.e., lysine and phenylalanine) had no influence on L[1-¹³C]phenylalanine oxidation when consuming the remaining amino acids at a constant protein intake equivalent to 1.0 g·kg⁻¹·d⁻¹ (33, 40). This was suggested to represent a strength of IAAO methodology to determine amino acid and/or protein requirements given the apparently short adaptation time required relative to the traditional nitrogen balance technique (i.e., ≤ 8 h vs. ≥5 days) (21). However, it was subsequently demonstrated that although the amount of protein (0.8, 1.4, or 2.0 g·kg⁻¹·d⁻¹) consumed in the 2 days prior to an IAAO metabolic trial (providing 1.0 g·kg⁻¹·d⁻¹) had no discernable pattern of influence on F¹³CO₂ there was significant variability in phenylalanine flux and oxidation in a small subset (n = 5) of healthy adults consuming a habitual intake of ~1.0–1.5 g protein·kg⁻¹·d⁻¹ (20). As a result, a 2-days adaptation period of controlled protein intake has been assumed to be sufficient prior to metabolic trials using IAAO methodology and have therefore been employed in studies estimating protein requirements in active populations who traditionally consume protein at intakes several fold that of the current RDA (5, 11, 41). In contrast to previous research, our F¹³CO₂ results suggest that high habitual protein consumers require at least 3 days and perhaps up to 5 days to adapt to a lower controlled dietary protein intake. These data would be consistent with an increased amino acid oxidative capacity with excess protein consumption (17, 18, 42) that in the case of the branched chain amino acids may be mediated by increased expression and/or activity of the rate controlling enzyme branched chain α-ketoacid dehydrogenase complex (43, 44). Therefore, our results demonstrate the traditional 2-days pre-trial controlled diet that is typical of most IAAO studies would be insufficient to ensure proper metabolic adaptation in populations who habitually consume high, and arguably excessive, protein intakes, which subsequently could result in a biased overestimation of true protein requirements.

The indicator amino acid is provided in excess of its metabolic requirement to ensure it is always partitioned toward oxidation, which would be lowest at a sufficient protein/amino acid intake. It is notable therefore that F¹³CO₂ (the most direct estimate of oxidation of the indicator amino acid) (21) at a surfeit protein intake (i.e., ≥2.2 g·kg⁻¹·d⁻¹) is similar between the present study and previous studies estimating the protein needs of high habitually consuming resistance-trained athletes (Table 2). The similar F¹³CO₂ observed in the present study compared to previous work (Table 2) suggests that an equivalent capacity to utilize the indicator amino acid to support protein synthesis (the reciprocal of F¹³CO₂) was obtained across the present and previous studies at these surfeit protein intakes. We also observed similar F¹³CO₂ on M3 (i.e., after a traditional 2-d dietary adaptation period) in the present study as the estimated F¹³CO₂ at 1.2 g·kg⁻¹·d⁻¹protein consumption in previous studies (11, 12). The similarity in the ratio of F¹³CO₂ at 1.2 and 2.2 g·kg⁻¹·d⁻¹ protein intake across studies (i.e., ~1.6–1.9) suggests the habitually high intakes of these populations had a similar impact on the ability of lower protein intakes to support protein synthesis. Although F¹³CO₂ was not statistically different after 5 days of adaptation to a moderate (1.2 g·kg⁻¹·d⁻¹) as compared to high (2.2.g·kg⁻¹·d⁻¹) protein intake it remained elevated by ~23%, which would be generally consistent with a requirement of up to 7 days to down-regulated leucine oxidation by ~35% after a ~1 g·kg⁻¹·d⁻¹ reduction in dietary protein (45). Taken together, the attenuation in F¹³CO₂ after initiation of a 2 days moderate 1.2 g·kg⁻¹·d⁻¹ protein intake could indicate that previous estimates of the protein requirements in strength-trained populations by IAAO may be overestimated by at least ~25% (i.e., M3 vs. M5) (12, 13) and potentially greater with higher [i.e., 1.5–1.6 g·kg⁻¹·d⁻¹ (11, 46)] adaptation intakes. Therefore, in our hands, an insufficient dietary protein adaptation period in high habitual protein consumers would have little impact on F¹³CO₂ at surfeit protein intakes but could result in a rightward shift in the bi-phase-determined breakpoint (i.e., EAR) and an overestimation of true protein requirements when using the IAAO.

Resistance training-induced increases in FFM can be supported by a range of dietary protein intakes (23). Data from infusion trials suggests post-exercise skeletal muscle anabolism can be maximized with 4–5 meals per day providing ~0.3 g·kg⁻¹ per meal or the equivalent (assuming a balanced distribution) of ~1.2–1.5 g·kg⁻¹·d⁻¹ (14, 16, 47). The present study demonstrates that reducing protein overconsumption in high habitual consumers results in an acute attenuation of whole-body net balance that approaches the anabolic potential of a high (excessive) dietary protein intake by ~5 days (M5). Importantly, phenylalanine flux was not influenced by protein intake suggesting there was no maladaptive down-regulation of whole-body protein metabolism during this acute adaptation phase. Thus, we argue that reductions in whole-body phenylalanine net balance primarily reflect an adaptation to amino acid oxidative capacity toward a more efficient dietary utilization of amino acids (i.e., a larger proportion used for protein synthesis vs. oxidation). Indeed, early adaptations (12 weeks) to resistance training in which muscle growth rate would be greatest (9) are supported by protein intakes of 1.2–1.4 g·kg⁻¹·d⁻¹ (24, 25), which would be substantially lower than the habitual consumption in this population (5). These findings suggest resistance trained men habitually consuming a high protein diet may be able to maintain whole body anabolism on a moderate protein intake provided adequate time to adapt to this lower intake. Conversely, the high habitual consumption in these populations would ultimately beget a reciprocally high metabolic demand for dietary amino acids to offset the oxidative losses, which may be reflected in the recent and arguably high estimated protein requirements of resistance trained populations by IAAO (11, 13). This increased metabolic demand for dietary amino acids would be consistent with the well-known diurnal cycling of body nitrogen (i.e., fasted state losses countered by fed state gains), which has a greater amplitude at high (i.e., ~2.1 g·kg⁻¹·d⁻¹) compared to moderate (i.e., ~1.6 g·kg⁻¹·d⁻¹) intakes (48, 49). In addition, higher protein intakes may also be needed to support the post-prandial utilization of dietary amino acids for muscle protein synthesis given that splanchnic extraction is greater with high protein diets while mixed and myofibrillar protein synthesis is attenuated after consumption of a moderate protein intake (42, 50). Therefore, future research should determine the impact of a reduction in habitual dietary protein intake on post-exercise rates of muscle protein synthesis in the fed state.

The present study utilized a relatively small sample size (n = 5) but one that is in line with seminar nitrogen balance research identifying protein requirements in active populations (n = 6–7/group) (7, 17, 51, 52) and previous IAAO method validation studies (n = 4–6) (20, 53, 54). This apparently small sample size could reduce statistical power and predispose our study to potential type II errors, which could be evident by the statistical trends we observed in some comparisons. To circumvent this, we also reported effect sizes for the comparisons, which revealed moderate to large effects for the statistical trends. Given the similar responses we observed across our population and the physiologically sensible changes in protein metabolism that would be consistent with a successful adaptation to a lower protein intake, we do not believe our conclusions would be altered by a greater sample size but rather statistical trends would merely reach the α = 0.05 threshold to be deemed significant.

In conclusion, our findings suggest that reducing dietary protein in high habitual protein consumers decreases whole body net balance acutely for at least 3 days and perhaps up to 5 days. Our data are consistent with previous data in sedentary individuals (19, 45) showing that, following an acute reduction in dietary protein intake, metabolic adaptations require ≥5 days to adapt to the new intake. Collectively, the 2 days adaptation period prior to metabolic trials that is typical of IAAO protocols may be inadequate when studying populations with high habitual protein intakes and lead to an erroneous overestimation of true protein requirements to support whole body protein synthesis. However, consistent with previous findings in novice weight lifters (24, 25), whole body anabolism may be supported by a moderate (1.2 g·kg⁻¹·d⁻¹) protein intake in otherwise high habitual protein consuming resistance trained men provided sufficient time is allotted to adapt to this lower intake.

The datasets generated for this study are available on request to the corresponding author.

The studies involving human participants were reviewed and approved by Health Sciences Research Ethics Board, University of Toronto. The patients/participants provided their written informed consent to participate in this study.

CT-G, DW, JG, and DM designed the research. CT-G, DW, and JM conducted the research. CT-G, DW, and JG analyzed the data. CT-G, DW, and DM wrote the paper with input from JM and JG. DM has primary responsibility for final content. All authors have read and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer KT declared a past collaboration with one of the authors DM to the handling editor.