Advances, Implications, and Limitations of Low-Crude-Protein Diets in Pig Production

Simple Summary Currently, five crystalline essential amino acids (lysine, methionine, threonine, tryptophan, and valine) are generally used, allowing animal nutritionists to formulate diets with low crude protein levels. Moreover, isoleucine may also be used depending on its economic value and the specific feeding program. Experimentally, it has been shown that further reduced crude protein levels can be achieved by supplemental histidine, leucine, and phenylalanine to the diets. However, decreasing the dietary crude protein level while maintaining optimal ratios of amino acids has shown contradictory effects on pigs’ growth performance. Due to the divergence in the literature and the importance for practical formulation strategies in the swine industry, a literature review and a meta-analysis were performed to estimate the minimum crude protein level that would not compromise pig performance. Based on the present review, there is a minimum crude protein level after which the growth performance of pigs can be compromised, even though diets are balanced for all essential amino acids. Considering average daily gain and the gain-to-feed ratio, respectively, these levels were estimated to be 18.4% and 18.3% crude protein for nursery, 16.1% and 16.3% crude protein for growing, and 11.6% and 11.4% crude protein for finishing pigs. Abstract Currently, five crystalline essential amino acids (Lys, Met, Thr, Trp, and Val) are generally used, allowing formulation of low-crude-protein (CP) diets. Moreover, Ile may also be used depending on its economic value and the specific feeding program. Experimentally, it has been shown that further reduced CP levels can be achieved by supplemental His, Leu, and Phe to the diets. However, decreasing the dietary CP level while maintaining optimal ratios of amino acids has shown contradictory effects on pigs’ growth performance. Due to the divergence in the literature and the importance for practical formulation strategies in the swine industry, a literature review and a meta-analysis were performed to estimate the minimum CP level that would not compromise pig performance. Based on the present review, there is a minimum CP level after which the growth performance of pigs can be compromised, even though diets are balanced for essential amino acids. Considering average daily gain and gain to feed, respectively, these levels were estimated to be 18.4% CP (95% confidence interval [CI]: 16.3 to 18.4) and 18.3% CP (95% CI: 17.4 to 19.2) for nursery, 16.1% CP (95% CI: 16.0 to 16.2) and 16.3% CP (95% CI: 14.5 to 18.0) for growing, and 11.6% CP (95% CI: 10.8 to 12.3) and 11.4% CP (95% CI: 10.3 to 12.5) for finishing pigs.

Additionally, we also estimated the higher levels of crystalline lysine that could be supplemented before compromising ADG and G:F, which were 0.42% (95% CI: 0.30-0.53) and 0.43% (95% CI: 0.35-0.50), (Figure 2A,B), respectively. Using an average of 0.425%, this is equivalent to 0.54% of L-lysine HCl (78.8% purity) or 0.71% of L-lysine sulfate (60.0% purity). To our knowledge, this is the first study to propose a maximum level of crystalline lysine supplementation and thus could be used as a starting point for future studies.
The use of a SID Lys:CP ratio may also prove to be a good estimator of to what extent CP could be lowered [6,53,54]. In this way, it was obtained that 6.6% (95% CI: 5.9-7.2) of SID Lys:CP is the breakpoint above which the ADG is compromised ( Figure 3A). When considering the G:F, the estimated level was the same, 6.6% (95% CI: 6.1-7.0) ( Figure 3B). These levels are quite similar to the 6.4% SIDLys:CP proposed by Millet et al. [6]. The use of the SID Lys:CP can be used as a reference to maintain sufficient EAA to meet the pigs' need for NEAA. The use of a SID Lys:CP ratio may also prove to be a good estimator of to what extent CP could be lowered [6,53,54]. In this way, it was obtained that 6.6% (95% CI: 5.9-7.2) of SID Lys:CP is the breakpoint above which the ADG is compromised ( Figure 3A). When considering the G:F, the estimated level was the same, 6.6% (95% CI: 6.1-7.0) ( Figure 3B). These levels are quite similar to the 6.4% SIDLys:CP proposed by Millet et al. [6]. The use of the SID Lys:CP can be used as a reference to maintain sufficient EAA to meet the pigs' need for NEAA.  Changes in ADG (A) and G:F (B) in response to the ratio between standard ilea digestible Lys to crude protein (SID Lys:CP) in the nursery phase using a broken-line analysis. The equation for ADG was ADG = 0.393 + 0.035 × zl (R 2 = 0.98) and the breakpoint was SID Lys:CP ratio at 6.6% (95% CI: 5.9-7.2) when ADG was 0.393 kg/d. The p-value for the asymptote was <0.001, for the slope it was <0.001, and for the breaking point it was <0.001. The equation for G:F was G:F = 0.682 + 0.041 × z1 (R 2 = 0.91) and breakpoint was SID Lys:CP ratio at 6.6% (95% CI: 6.1-7.0 ) when G:F was 0.682. The p-value for the asymptote was <0.001, for the slope it was <0.001, and for the breaking point it was <0.001. Where, if SID Lys:CP is ≤breakpoint, then z1 = 0; if SID Lys:CP is >breakpoint, then z1 = SID Lys:CP − breakpoint [6,9,11,19,23,24,26,27,29,30,[35][36][37][49][50][51][52].
In summary, low-CP diets should be formulated by assuming a minimum CP level to avoid the limitation of other nutrients that may be deficient when reducing the CP below a certain level. Based on the present review, the minimum CP level is 18.3% for nursery diets with the concentration of EAA being maintained as recommended (NRC, 2012). Moreover, 0.54% of L-lysine HCl or 0.71% of L-lysine sulfate were estimated to be the maximum supplementation levels above which growth performance might be compromised. Finally, 6.6% SID Lys:CP was estimated to be the breakpoint for growth performance.

Growing Phase
Compared to the nursery, growing phase diets are formulated to a lower CP level. At the beginning of the growing phase, a corn-SBM diet with 17.5% CP could be formulated, supplementing only Lys, Met, Tre, and Trp to achieve the suggested amino acid requirements [1,3]. Lower CP levels can be obtained with the additional supplementation of the next limiting amino acids.
Madrid et al. [55] demonstrated that diets formulated using the ideal protein concept and supplemented with amino acids did not affect performance when dietary CP decreased from 16.0 to 14.0%. Supplementing all the EAA except Leu, Qiu et al. [56] demonstrated that CP could be decreased from 18.0 to 14.0% without compromising growth performance. According to Powell et al. [57], the growth performance of pigs was maintained when the dietary CP level was reduced from 18.2 to 13.4% while keeping the proper ratio of amino acids. In agreement, Zhao et al. [14] also found that a reduction in dietary CP from 17.4 to 13.5% with all EAAs balanced had no negative effect on pig growth performance.
Contrary to the previous studies, Li et al. [58], showed that decreasing the CP level from 18.3% to 15.1% resulted in decreased ADG and G:F although the calculated ratio of all EAA to Lys was in agreement with the NRC [1]. In this way, others have balanced diets for all the EAA but also failed to lower CP while maintaining similar growth performance. For example, Peng et al. [11] showed that pig growth performance was similar when dietary CP level was reduced from 20.0 to 15.3%; however, a further decrease to 13.9% resulted in lower ADG and G:F compared with pigs receiving a 20.0% CP diet. According to Che et al. [8], a moderate reduction in CP from 16.7 to 14.7% maintained the growth performance of pigs; however, a reduction to 12.9% markedly decreased growth performance. Additionally, Roux et al. [59] also failed to show similar growth performance with the supplementation of EAA in low-CP diets (18.2 vs. 13.4%).
The discrepancies among studies could be due to the variation in CP levels designed for the positive control diets [8]. A dietary CP level of 16.0 or 20.0% as a positive control are different concerning intact protein, peptides, and NEAA content. In this way, a further reduction of CP level may differentially affect animal growth performance. Similarly, Peng et al. [11] and Lynegaard et al. [46] also report that their positive control diets were already a lower CP level for pigs, and therefore, the lower-CP treatments may have compromised the growth performance of growing pigs. Moreover, it has been suggested that the energy content of SBM proposed in the tables [1,3] might be underestimated [60,61]. Thus, the ME of SBM can also affect the results of studies aiming at lower CP levels.
Similar to the nursery phase, the extent to which the CP content of the diets can be reduced without affecting the growth performance has been conflicting in the growing phase. Based on the model from the meta-analysis, CP content can be decreased to 16.1% (95% CI: 16.0-16.2) and 16.3% (95% CI: 14.5-18.0) when formulating for optimal ADG and G:F, as long as EAA meet the requirements of pigs ( Figure 4A,B). For a similar phase, the NRC [1] and Brazilian Tables [3] suggest 15.7% and 18.9% CP, respectively. The discrepancies among CP levels could be due to the variation in genotypes as the efficiency of utilization of dietary protein for body protein deposition is strongly related to pig genotype [1,3]. The present results are in between those suggestions. However, when below those levels of CP, other factors should be considered, such as the level of other nutrients (NEAA, peptides, and others).
For the growing phase, it was not possible to estimate a breakpoint for L-lysine supplementation on growth performance. However, the average maximum level of L-lysine supplementation in the growing trials was 0.42 ± 0.14% (Table 2), which is equivalent to 0.53% of L-lysine HCl (78,8% purity) or 0.70% of L-lysine sulfate (60% purity). This maximum level of feed-grade lysine supplementation is quite similar to the level estimated for the nursery phase. Over-feeding crystalline lysine and other amino acids can be costly, and therefore, the knowledge of a maximum level above which performance is compromised is important. [3] suggest 15.7% and 18.9% CP, respectively. The discrepancies among CP levels could be due to the variation in genotypes as the efficiency of utilization of dietary protein for body protein deposition is strongly related to pig genotype [1,3]. The present results are in between those suggestions. However, when below those levels of CP, other factors should be considered, such as the level of other nutrients (NEAA, peptides, and others). 2) when ADG was 0.787 kg/d. The p-value for the asymptote was <0.001, for the slope it was 0.050, and for the breaking point it was <0.001. The equation for G:F was G:F = 0.482 -0.008 × z1 (R2 = 0.93) and breakpoint was CP level at 16.3% (95% CI: 14.5-18.0) when G:F was 0.482. The p-value for the asymptote was <0.001, for the slope it was 0.008, and for the breaking point it was <0.001. Where, if CP is ≥breakpoint, then z1 = 0; if CP is <breakpoint, then z1 = CP -breakpoint. [7,8,11,13,14,18,[55][56][57][58][59][62][63][64][65][66][67]. Considering the studies in the meta-analysis, it was not possible to estimate a breakpoint for SID Lys:CP ratio on growth performance of growing pigs. As mentioned before, this is a relatively new estimator, recently proposed by Millet et al. [6] working with nursery pigs. 1 Supplemented feed grade L-lysine based on the purity of L-lysine HCl or L-lysine sulfate in the lower-CP diet. 2 Additional supplemented feed grade amino acids in low-CP diets in addition to Lys, Met, Tre, and Trp. 3 References listed in year order, and only studies and treatments balanced for all essential amino acids were used. When necessary, amino acids were estimated after reformulating the diets [1]. * Estimated levels.

Finishing Phase
Protein is a relatively expensive nutrient and due to the increase in feed intake during the finishing phases, the efficiency of nutrient use greatly impacts the costs incurred by the production system [4]. An early-finishing (70 to 100 kg) diet may be formulated with approximately 12.3% CP and a late-finishing (100 kg to slaughter) diet with approximately 11.0% CP while maintaining the concentration of EAA as recommended [1,3].
In agreement, Xie et al. [68] and Zhou et al. [69] found that the CP levels could be reduced from 15.3 to 12.0% and from 14.1 to 10.1% CP, respectively, without affecting growth performance in the early-finishing phase. In addition to similar performance, Norgaard et al. [70] observed that a reduction in CP from 15.9 to 13.6% did not affect carcass characteristics. Qin et al. [17] also demonstrated no effect on the growth performance, carcass characteristics, and meat quality of late-finishing pigs when CP was reduced from 14.3 to 12.3%.
However, some studies evaluating low-CP diets have reported decreases in growth performance, even when diets are balanced for the limiting EAA. Li et al. [71] showed that decreasing the CP level from 16.3% to 13.2% resulted in decreased ADG and G:F. It is interesting and worthwhile to note that the lower-CP diet improved meat quality. Soto et al. [12] evaluated the effect of CP levels for early-(from 13.1 to 9.0% CP) and late-(from 12.9 to 8.9% CP) finishing pigs. They observed that reduction in CP negatively affected growth performance in both trials. However, carcass yield, backfat, and loin depth were not affected by CP levels. Comparing 16.0 and 12.0% CP diets, Zhou et al. [72] showed that pigs fed a low-CP diet had lower ADG and G:F, with no effect on body fat content.
Considering that all EAA were properly supplied in the cited papers, there are some explanations in the literature for the negative effects on growth when low-CP diets are fed. As previously mentioned, a deficiency of NEAA in low-CP diets may affect pig growth performance. Moreover, supplementing EAA leads to a reduction in the supplementation of protein sources, such as soybean meal. These ingredients contain biologically active compounds, such as isoflavones, saponins, and bioactive peptides, that may also be important to maintain growth performance [10,12,44]. Diet-derived peptides can exert actions at the level of the small intestine, with the resulting intestinally-generated signals having impacts on the whole body. Bioactive peptides have been suggested to possess antimicrobial, antioxidant, and immunomodulatory activities [73]. Furthermore, it is noteworthy that a significant amount of small peptides are absorbed intact [74,75], and the specificity of protein hydrolysis and the related release of peptides differ between feed ingredients, causing different effects on animal physiology [76]. Thus, reducing the protein sources to achieve a lower CP level may affect animal growth performance.
Apart from improvements in growth performance and carcass characteristics, the swine feed industry has been pushed to reduce dietary CP and simultaneously the excretion of nitrogen, which can contribute to greenhouse gas emissions [14,17,18]. In this way, it is well established the benefits of reduced CP diets on lowering the environmental impact of pig production [14,17,70]. However, the relationship between low CP and the environment was not the focus of the present review and may be found elsewhere [18,77]. Meanwhile, new research should continue to focus on strategies utilizing low-CP diets while not compromising growth performance and pig profitability.
Similar to the finishing phase, contradictory results have been reported when considering the entire growing to finishing phases. Working with growing-finishing pigs (25-110 kg) in two climatic conditions, Wang et al. [66] evaluated the effect of low and high CP diets. In both experiments, pigs fed low-CP diets supplemented with EAA had similar growth performance, carcass characteristics, and meat quality to pigs fed high CP diets. In the same way, Zhao et al. [14] showed no effects of decreasing CP in the diets on growth performance and carcass characteristics of growing-finishing pigs (25-125 kg). Recently, Le Dinh et al. [77] found that reducing CP with EAA supplementation increased ADG and ADFI but also reduced muscle thickness. Conversely, Fang et al. [65] found that growing/finishing pigs (30-105 kg) fed low-CP diets supplemented with EAA resulted in reduced growth performance.
In agreement, the model from the meta-analysis suggested that the CP level below 11.6% (95% CI: 10.8-12.3) and 11.4% (95% CI: 10.3-12.5) can compromise the ADG and G:F, respectively ( Figure 5). These levels are in between the levels suggested in NRC [1] for early-(12.1% CP) and late-(10.5% CP) finishing pigs, probably because we considered both phases together in our model. On the other side, the proposed levels are below the average recommendations in Brazilian Tables [3] for early-(13.3% CP) and late-(11.9% CP) finishing pigs. Thus, CP levels lower than the Brazilian Tables [3] recommendations might be used without compromising pig growth performance. G:F, respectively ( Figure 5). These levels are in between the levels suggested in NRC [1] for early-(12.1% CP) and late-(10.5% CP) finishing pigs, probably because we considered both phases together in our model. On the other side, the proposed levels are below the average recommendations in Brazilian Tables [3] for early-(13.3% CP) and late-(11.9% CP) finishing pigs. Thus, CP levels lower than the Brazilian Tables [3] recommendations might be used without compromising pig growth performance. 3) when ADG was 0.924 kg/d. The p-value for the asymptote was <0.001, for the slope it was 0.050, and for the breaking point it was <0.001. The equation for G:F was G:F = 0.327 -0.014 × z1 (R 2 = 0.96) and breakpoint was CP level at 11.4% (95% CI: 10.3-12.5) when G:F was 0.327. The p-value for the asymptote was <0.001, for the slope it was 0.006, and for the breaking point it was <0.001. Where, if CP is ≥breakpoint, then z1 = 0; if CP is <breakpoint, then z1 = CP -breakpoint. [12,17,68,69,[70][71][72][78][79][80][81][82].
Concerning feed grade lysine, the breakpoint for ADG was 0.24% L-lysine (95% CI: 0.10-0.37; Figure 6), which is equivalent to the supplementation of 0.30% L-lysine-HCl or 0.40% L-lysine sulfate. The proposed level of L-lysine is lower than that for nursery and growing phases because lysine supplementation in finishing diets is notably lower than in the initial phases (Table 3). Considering the studies in the meta-analysis, a breakpoint for G:F was not found (a model was not fit). Additionally, a breakpoint for SIDLys:CP ratio on growth performance was not found. 0.40% L-lysine sulfate. The proposed level of L-lysine is lower than that for nursery and growing phases because lysine supplementation in finishing diets is notably lower than in the initial phases (Table 3). Considering the studies in the meta-analysis, a breakpoint for G:F was not found (a model was not fit). Additionally, a breakpoint for SIDLys:CP ratio on growth performance was not found. Figure 6. Changes in ADG in response to L-lysine supplementation in the finishing phase using a broken-line analysis. The L-lysine axis is based on 100% purity, calculated from L-lysine HCl (78.8% purity) or L-lysine sulfate (60.0% purity). The equation was ADG = 0.926 + 0.314 × zl (R 2 = 0.97) and the breakpoint was L-lysine level at 0.24% (95% CI: 0.10-0.37) when ADG was 0.926 kg/d. The pvalue for the asymptote was <0.001, for the slope it was 0.033, and for the breaking point it was 0.002. Where, if L-lysine is ≤breakpoint, then z1 = 0; if L-lysine is >breakpoint, then z1 = L-lysinebreakpoint. [12,17,68,69,[70][71][72][78][79][80][81][82].

Conclusions
In conclusion, there is a minimum CP level after which the growth performance of pigs can be compromised, even though diets are balanced for all EAA. Apparently, there is a level after which other nutrients such as NEAA, bioactive compounds, and others become limiting. Based on the meta-analysis, considering the ADG and G:F the minimum levels of CP were estimated at 18.4% and 18.3% for nursery, 16.1% and 16.3% for growing, and 11.6% and 11.4% for finishing pigs, respectively. Moreover, it was estimated the higher levels of L-lysine (100% purity) to be supplemented before compromising growth performance were 0.42% for ADG and 0.43% for G:F in the nursery phase and 0.24% for ADG in the finishing phase. Additionally, a level of 0.42% L-lysine supplementation is suggested for growing pigs. Finally, it was obtained that 6.6% of SID Lys:CP is the ratio above which the ADG and G:F of nursery pigs are compromised. Considering that feed grade L-lysine HCl is 78% lysine, based on the meta-analysis, optimal allowances of the use of L-lysine HCl are 0.54% in feeds for nursery and growing pigs and 0.31% in feeds for finishing pigs. Applying these optimal levels of L-lysine HCl in swine feeds should not compromise the growth performance of pigs if the ideal ratios of lysine to other EAA and ME are considered.  Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: Publicly available datasets were analyzed in this study. This data can be found here: https://www.ncbi.nlm.nih.gov/.