Camp Montoro et al. Porcine Health Management (2022) 8:32 https://doi.org/10.1186/s40813-022-00273-y RESEARCH Open Access Blood and faecal biomarkers to assess dietary energy, protein and amino acid efficiency of utilization by growing and finishing pigs Jordi Camp Montoro1,2*, David Solà‑Oriol2, Ramon Muns3, Josep Gasa2, Núria Llanes4 and Edgar Garcia Manzanilla1,5 Abstract Background: Diet evaluation and optimization is a slow and expensive process and it is not possible to do it at a farm level. This study aimed to use the blood serum metabolite (BSM) and faecal volatile fatty acid (VFA) profiles as potential biomarkers to identify changes in protein, amino acid and energy dietary content in growing and finishing pig diets at farm level. Results: Two studies were conducted. The first study (S1) included 20 pens of 11 pigs (87.0 ± 4.10 kg; 18 weeks old) assigned to 5 diets: control (C1), high or low crude protein (HP1 and LP1, respectively), and high or low net energy (HE1 and LE1, respectively). The second study (S2) included 28 pens of 11 pigs (41.3 ± 2.60 kg; 12 weeks old) assigned to 7 diets: control (C2), high or low crude protein (HP2 and LP2, respectively), high or low amino acid (HA2 and LA2, respectively), and high or low net energy (HE2 and LE2, respectively). Pigs were followed for 10 (S1) and 20 (S2) days, and blood and faecal samples were collected at 20 (S1) and 14 (S2) weeks of age. Data were analysed using gen‑ eral linear models and receiver operating characteristic curve analysis. Urea nitrogen showed the best results as a biomarker. Urea nitrogen was higher in pigs fed high protein diets, HP1 (13.6 ± 0.95 mg/dL) and HP2 (11.6 ± 0.61), compared to those fed low protein diets, LP1 (6.0 ± 0.95) and LP2 (5.2 ± 0.61; P < 0.001), showing good discrimination ability (Area under the curve (AUC) = 98.4 and 100%, respectively). These differences were not observed between diets LA2 (6.5 ± 0.61) and HA2 (8.7 ± 0.61; P > 0.05; AUC = 71.9%), which were formulated based on the ideal protein profile but with no excess of protein. Creatinine, triglycerides, branched-chain fatty acids, albumin, propionic acid, and cholesterol showed differences between at least 2 diets but only in one of the studies. Conclusions: Urea nitrogen showed high accuracy to detect excess of crude protein in growing and finishing pig diets. Other biomarkers like BCFA showed promising results and need to be further studied. Keywords: Faeces, Growth, Lysine, Metabolism, Nutrition, Requirements, Swine Background Pig diets are formulated with the goal of optimising *Correspondence: jordi.camp95@gmail.com growth, health, and welfare of the animals by meeting 1 their nutritional requirements while avoiding the excess Pig Development Department, Animal and Grassland Research and Innovation Centre, Teagasc, Moorepark, Fermoy, Co. Cork P61 C996, of nutrients, especially nitrogen and phosphorus, that Ireland result in environmental pollution. However, the actual Full list of author information is available at the end of the article © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​ mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Camp Montoro et al. Porcine Health Management (2022) 8:32 Page 2 of 14 nutritional value of a diet for a particular pig farm is evaluate the use of the blood serum metabolite and fae- affected by different factors such as quality of ingredients cal VFA profiles as potential biomarkers to identify dif- [1], feed form and delivery method [2], variability of body ferences in the levels of dietary net energy (NE), crude weight (BW) [3, 4], farm management [5, 6], or health protein (CP), and AA within normal ranges for growing status of the farm [7, 8], among others. With so many fac- and finishing pigs at farm level. tors affecting nutritional value of feed, suboptimal diets are not rare and can result in extra cost for the farmers, Methods potential health and welfare problems for the pigs, and Animals, diets, and experimental design environmental contamination. Two studies were conducted in the Teagasc Pig Research The use of low protein diets, including more synthetic Facility in Fermoy, Co. Cork, Ireland. Both studies amino acids (AAs), is one of the most effective measures received ethical approval from the Teagasc Animal Eth- to reduce the incidence of diarrhoea in pigs [9] and to ics Committee (TAEC 244/2019). In both studies, pigs reduce ammonia emissions in pig farms [10]. However, were housed in pens with fully slatted concrete floor the optimum levels of AAs and the balance between AAs (2.4 × 4.2 m) containing a single wet-dry feeder (330 mm and energy can vary depending on the particularities of [Width] × 370 mm [Depth] × 1000 mm [Height]; MA37, each farm, and the same low protein diet can be adequate Verba, Netherlands) and one supplementary nipple or not depending on the farm. Thus, methods to opti- drinker. Water and pelleted feed were provided ad libi- mise diets at farm level are needed. Classical methods for tum. Temperature was controlled by a mechanical venti- feed assessment, such as digestibility trials, are expensive, lation system with fan speed and air inlet area regulated time consuming, and are not suited for use in commer- by a climate controller. Pens were enriched with a larch cial farms [11]. Fast methods to optimise diets at individ- wood post. ual farm level would be of interest to improve production The first study (S1) was conducted in finishing pigs efficiency, animal health and welfare, and to reduce the from 18 to 20 weeks of age (87.0 ± 4.10 to 100.3 ± 4.40 kg environmental footprint and production costs [12–14]. of BW). A total of 220 Danish Duroc × (Large Blood biochemistry is a fast analysis method used White × Landrace) growing pigs born within one week regularly in clinical practice for many animal species were moved to the grower-finisher stage at 11 weeks of and it can provide parameters directly related to energy age and housed in balanced mixed sex pens. Pigs were and protein metabolism. Dietary changes affect blood fed a single soybean meal-maize-wheat based grower- metabolites such as total protein [15–17], albumin [16– finisher diet (9.67 MJ/kg of NE, 16.2% of CP, and 0.92% 18], serum urea nitrogen (SUN) [19–21] and creatinine of standard ileal digestive (SID) lysine (Lys)) from 11 [21] for the protein metabolism, and glucose [15–17], tri- to 18 weeks of age before the study started. The study glycerides [18, 22] and cholesterol [17, 18] for the energy started at 18 weeks of age when pigs were weighed per metabolism. pen (n = 20; 11 pigs/pen) and assigned per pen based Faecal samples are easy to collect in pig farms and on BW to five different dietary treatments following a fermentation components present in faeces, like vola- complete randomized design. Pigs were followed for tile fatty acids (VFAs) can be easily measured and are 10 days. Diets were formulated to obtain a control diet directly affected by the composition of the diet and the (C1; 10.03 MJ/kg of NE, 16.0% of CP, and 0.95% of SID metabolism of nutrients by the animal [23–25]. Carbo- Lys) which met or exceed the minimum nutrient require- hydrates are catabolized to short-chain fatty acids such ments [26], and 4 modifications of this diet: low crude as acetic, propionic or butyric acid, while protein results protein (LP1) by reducing CP to 13.2% and AAs to 0.75% in a higher concentration of branched-chain fatty acids of SID Lys; high crude protein (HP1) by increasing CP (BCFA) produced from the deamination of branched- to 18.8% and AAs to 1.15% of SID Lys; low net energy chain amino acids such as leucine, isoleucine, and valine (LE1) by reducing NE to 9.61 MJ/kg; and high net energy [25]. Therefore, high proportions of BCFA measured in (HE1) by increasing NE to 10.45 MJ/kg. Ingredient, cal- faeces could indicate an excess of protein reaching distal culated, and analysed nutrient diet composition is shown parts of the intestine and being inefficiently used by pigs. in Table 1. Pigs went back to the common management Blood serum metabolites and faecal VFA profiles could of the Teagasc Pig Research Facility after the 10-day trial be potential biomarkers to identify diets that are subop- period. timal at individual farm level. Therefore, we hypothesise The second study (S2) was conducted in growing pigs that blood metabolites and faecal VFA profiles are bio- from 12 to 15 weeks of age (41.3 ± 2.60 to 62.1 ± 3.66 kg markers sensible enough to reflect changes in dietary of BW). A total of 308 Danish Duroc × (Large energy, protein, and AA for growing and finishing pigs White × Landrace) growing pigs born within one week at farm level. The objective of the present study was to were moved to the grower-finisher stage at 11 weeks of Camp Montoro et al. Porcine Health Management (2022) 8:32 Page 3 of 14 Table 1 Ingredient, calculated, and analysed nutrient Table 1 (continued) composition on an as-fed basis of the five dietary treatments in Dietsb study ­1a C1 LP1 HP1 LE1 HE1 Dietsb Digestible phosphorus 0.22 0.22 0.22 0.25 0.22 C1 LP1 HP1 LE1 HE1 a Diets were fed in finishing pigs during 10 days at 18 weeks of age b Ingredients, g/kg C1 = Control; LP1 = Low Crude Protein; HP1 = High Crude Protein; LE1 = Low Net Energy; HE1 = High Net Energy Wheat 350.0 350.0 350.0 330.0 306.2 c Provided per each kg of feed: 60 mg Copper sulphate, 80 mg Ferrous sulphate Barley 282.5 345.0 0.0 310.5 200.0 monohydrate, 50 mg Manganese oxide, 100 mg Zinc oxide, 0.5 mg Potassium Maize 150.0 150.0 286.6 100.0 275.5 iodate, 0.4 mg Sodium selenite, 2 MIU Vitamin A, 0.5 MIU Vitamin ­D3, 40 MIU Vitamin E, 4 mg Vitamin K, 0.015 mg Vitamin ­B12, 2 mg Riboflavin, 12 mg Soybean meal 47.5 172.4 95.7 254.1 175.1 172.4 Nicotinic acid, 10 mg Pantothenic acid, 2 mg Vitamin ­B1, 3 mg Vitamin ­B6 Soybean hulls 14.2 29.7 63.9 58.3 0.0 d NE = Net Energy; SID = Standardized Ileal Digestible; NDF = Neutral Detergent Vegetable oil 5.0 5.0 17.6 0.0 21.5 Fiber Calcium carbonate 12.3 12.7 12.2 10.7 11.7 Dicalcium phosphate anhydrous 0.50 0.50 1.00 3.00 0.50 age and housed in balanced mixed sex pens. After one Sodium chloride 4.50 4.40 3.20 4.40 3.70 week of adaptation, pigs were weighed per pen (n = 28 L-Lysine HCl 4.30 3.80 5.30 4.15 4.40 pens; 11 pigs/pen) and assigned per pen based on BW L-Threonine 1.60 1.20 2.20 1.15 1.60 to seven different dietary treatments at 12 weeks of age, DL-Methionine 1.30 0.70 2.20 1.30 1.20 following a complete randomized design. Pigs were fol- L-Tryptophan 0.20 0.10 0.20 0.20 0.10 lowed for 20 days. Diets were formulated to obtain a con- L-Valine 0.00 0.00 0.30 0.00 0.00 trol diet (C2; 10.03 MJ/kg of NE, 16.5% of CP, and 0.95% Vitamin and trace mineral 1.20 1.20 1.20 1.20 1.20 ­mixturec of SID Lys) which met or exceed the minimum nutrient Calculated/analysed ­compositiond, % as fed or as specified requirements [26], and 6 modifications of this diet: low Dry matter, analysed 88.00 87.70 88.30 87.90 87.90 crude protein (LP2) by reducing CP to 14.0%; high crude Ash, analysed 3.90 3.60 4.00 4.10 3.90 protein (HP2) by increasing CP to 19.0%; low amino acid NE, MJ/kg 10.03 10.03 10.03 9.61 10.45 (LA2) by reducing AAs to 0.75% of SID Lys; high amino SID Lys:NE, g/MJ 0.95 0.75 1.15 0.99 0.91 acid (HA2) by increasing AAs to 1.15% of SID Lys; low Crude Protein, analysed 13.40 11.60 16.20 14.50 14.30 net energy (LE2) by reducing NE to 9.61 MJ/kg; and Total Lys, analysed 1.05 0.88 1.31 1.08 1.02 high net energy (HE2) by increasing NE to 10.45 MJ/kg. Total Thr/Lys ratio, analysed 0.58 0.58 0.56 0.57 0.62 When the level of amino acids was modified all amino Total Met-Cys/Lys ratio, analysed 0.64 0.64 0.63 0.65 0.68 acids were adjusted following the ideal protein concept Total Trp/Lys ratio, analysed 0.14 0.15 0.15 0.14 0.14 [26, 27]. Ingredient, calculated, and analysed nutrient Total Val/Lys ratio, analysed 0.60 0.65 0.65 0.67 0.69 diet composition is shown in Table 2. Pigs went back to Total Leu/Lys ratio, analysed 1.09 1.14 1.08 1.14 1.14 the common management of the Teagasc Pig Research Total Ile/Lys ratio, analysed 0.54 0.55 0.57 0.58 0.60 Facility after the 20-day trial period. Total His/Lys ratio, analysed 0.34 0.38 0.37 0.36 0.38 Total Phe/Lys ratio, analysed 0.61 0.64 0.65 0.69 0.69 Body weight, feed intake, and feed efficiency traits Total Tyr/Lys ratio, analysed 0.25 0.25 0.30 0.32 0.30 On both studies, S1 and S2, pigs were weighed per pen at Total Arg/Lys ratio, analysed 0.81 0.82 0.86 0.89 0.88 the beginning and at the end of the trial period. Feed bags SID Lys 0.95 0.75 1.15 0.95 0.95 were used to feed the pigs. Overall feed intake was SID Thr/Lys ratio 0.65 0.65 0.65 0.65 0.65 recorded per pen by weighing the feed bags before feed- SID Met-Cys/Lys ratio 0.59 0.59 0.59 0.59 0.59 ing the pigs and before weighing the pigs per pen (includ- SID Trp/Lys ratio 0.19 0.19 0.19 0.19 0.19 ing the feed inside the feeder) at the end of the trial. SID Val/Lys ratio 0.66 0.67 0.65 0.66 0.66 Average daily gain (ADG), average daily feed intake SID Leu/Lys ratio 1.16 1.20 1.11 1.16 1.15 (ADFI) and feed conversion ratio (FCR) were calculated SID Ile/Lys ratio 0.56 0.55 0.57 0.57 0.57 for the overall trial period. Feed conversion ratio was cal- SID His/Lys ratio 0.36 0.36 0.34 0.35 0.36 culated as kgoffeedconsumed BWgain . Fat, analysed 2.79 2.74 3.78 2.21 4.19 Crude fiber, analysed 2.90 3.40 4.20 4.20 2.40 Feed analysis NDF 12.96 14.15 13.54 15.02 12.00 Feed samples of each diet were collected per duplicate Calcium 0.75 0.75 0.80 0.77 0.72 and analysed for dry matter, ash, CP, crude fibre, fat, and total AA profile at the Sciantec Analytical Services Camp Montoro et al. Porcine Health Management (2022) 8:32 Page 4 of 14 Table 2 Ingredient, calculated, and analysed nutrient composition on an as-fed basis of the seven dietary treatments in study 2­ a Dietsb C2 LP2 HP2 LA2 HA2 LE2 HE2 Ingredients, g/kg Maize 272.9 401.1 254.9 350.0 423.1 300.0 332.1 Barley 251.8 230.0 252.2 295.9 150.0 246.1 225.7 Wheat 218.1 150.0 150.0 150.0 150.0 150.0 150.0 Soybean meal 47.5 209.3 137.2 286.7 135.9 197.6 209.8 218.7 Soybean hulls 20.0 44.0 20.0 38.9 39.8 70.4 20.0 Vegetable oil 5.00 5.00 17.6 5.00 5.00 0.00 31.0 Calcium carbonate 10.6 14.0 10.5 13.6 12.3 11.3 10.6 Dicalcium phosphate anhydrous 0.50 0.50 0.50 0.50 0.70 0.50 0.50 Sodium chloride 4.80 5.00 4.30 4.50 5.20 4.90 4.50 L-Lysine HCl 3.00 5.20 0.80 2.60 6.00 2.90 2.90 L-Threonine 1.30 2.30 0.40 1.00 2.80 1.30 1.30 DL-Methionine 1.20 1.80 0.70 0.60 2.50 1.30 1.20 L-Tryptophan 0.10 0.50 0.00 0.10 0.60 0.10 0.10 L-Valine 0.00 0.90 0.00 0.00 1.30 0.00 0.00 L-Isoleucine 0.00 0.90 0.00 0.00 0.70 0.00 0.00 L-Leucine 0.00 0.00 0.00 0.00 0.60 0.00 0.00 L-Histidine 0.00 0.20 0.00 0.00 0.40 0.00 0.00 ­ ixturec Vitamin and trace mineral m 1.40 1.40 1.40 1.40 1.40 1.40 1.40 Calculated/analysed ­compositiond, % as fed or as specified Dry matter, analysed 87.30 87.70 88.10 87.70 88.00 87.50 88.30 Ash, analysed 3.60 4.10 4.20 4.10 4.20 4.10 3.80 NE, MJ/kg 10.03 10.07 10.03 10.03 10.07 9.61 10.45 SID Lys:NE, g/MJ 0.95 0.94 0.95 0.75 1.14 0.99 0.91 Crude protein, analysed 15.4 15.1 17.7 13.7 16.6 16.2 16.3 Total Lys, analysed 1.02 1.12 1.08 0.86 1.32 1.07 1.02 Total Thr/Lys ratio, analysed 0.71 0.69 0.73 0.74 0.67 0.68 0.72 Total Met-Cys/Lys ratio, analysed 0.62 0.62 0.63 0.64 0.58 0.60 0.63 Total Trp/Lys ratio, analysed 0.19 0.15 0.18 0.09 0.15 0.16 0.17 Total Val/Lys ratio, analysed 0.75 0.68 0.87 0.76 0.67 0.74 0.75 Total Leu/Lys ratio, analysed 1.19 0.98 1.38 1.27 1.04 1.18 1.23 Total Ile/Lys ratio, analysed 0.63 0.56 0.76 0.64 0.57 0.63 0.64 Total His/Lys ratio, analysed 0.41 0.37 0.48 0.40 0.37 0.41 0.42 Total Phe/Lys ratio, analysed 0.70 0.55 0.82 0.73 0.57 0.64 0.69 Total Tyr/Lys ratio, analysed 0.29 0.23 0.28 0.23 0.27 0.25 0.25 Total Arg/Lys ratio, analysed 0.95 0.74 1.16 0.96 0.78 0.93 0.98 SID Lys 0.95 0.95 0.95 0.75 1.15 0.95 0.95 SID Thr/Lys ratio 0.65 0.65 0.65 0.65 0.65 0.65 0.65 SID Met-Cys/Lys ratio 0.59 0.59 0.59 0.59 0.59 0.59 0.59 SID Trp/Lys ratio 0.19 0.19 0.22 0.19 0.19 0.19 0.19 SID Val/Lys ratio 0.69 0.65 0.81 0.72 0.65 0.69 0.69 SID Leu/Lys ratio 1.14 0.98 1.33 1.23 0.99 1.15 1.16 SID Ile/Lys ratio 0.59 0.55 0.71 0.60 0.52 0.59 0.60 SID His/Lys ratio 0.37 0.32 0.43 0.38 0.32 0.36 0.37 Fat, analysed 2.28 2.75 3.91 2.64 2.89 2.40 5.00 Crude fiber, analysed 2.90 3.50 3.20 3.70 3.20 4.80 2.90 NDF 12.52 13.72 12.00 14.00 12.72 14.81 12.00 Calcium 0.70 0.80 0.70 0.80 0.77 0.75 0.70 Digestible phosphorus 0.22 0.22 0.22 0.22 0.22 0.22 0.22 Camp Montoro et al. Porcine Health Management (2022) 8:32 Page 5 of 14 Table 2 (continued) a Diets were fed in growing pigs during 20 days at 12 weeks of age b C2 = Control; LP2 = Low Crude Protein; HP2 = High Crude Protein; LA2 = Low Amino Acid; HA2 = High Amino Acid; LE2 = Low Net Energy; HE2 = High Net Energy c Provided per each kg of feed: 60 mg Copper sulphate, 80 mg Ferrous sulphate monohydrate, 50 mg Manganese oxide, 100 mg Zinc oxide, 0.5 mg Potassium iodate, 0.4 mg Sodium selenite, 2 MIU Vitamin A, 0.5 MIU Vitamin ­D3, 40 MIU Vitamin E, 4 mg Vitamin K, 0.015 mg Vitamin ­B12, 2 mg Riboflavin, 12 mg Nicotinic acid, 10 mg Pantothenic acid, 2 mg Vitamin ­B1, 3 mg Vitamin ­B6 d NE = Net Energy; SID = Standardized Ileal Digestible; NDF = Neutral Detergent Fiber (Stockbridge Technology Centre, Cawood, Yorkshire, and S2 respectively. BioFreeze™ vials enable to collect the (n = 8 per treatment) faecal samples were collected in S1 UK). Dry matter was measured by oven drying for 4 h at 103 °C [28]; ash was measured via combustion in a muf- fresh samples and stop all biological activity at ambient fle furnace at 550 °C [29]; CP was determined as N × 6.25 temperature until the analysis. Faecal VFA analysis was based on the DUMAS method [30] using LECO FP-628 conducted via gas chromatography using pivalic acid as analyser (Leco Instruments Ltd., Stockport, UK); crude an internal standard [33] at Alimetrics Diagnostics. The fibre was determined by a Fibertec semi-automatic sys- VFA profile included acetic, propionic, butyric, valeric, tem (Tecator, Hoganas, Sweden) using the gravimetry BCFA and total VFA. method [28]; fat was measured using Randall/Soxtec/ Submersion method [31]; and total AA profile was deter- Data management and statistical analysis mined based on ion exchange HPLC [32] using the Bio- Growth performance, blood serum and faecal VFA data chrom AA Analyser Sodium System (Biochrom Ltd., analyses were carried out using SAS v9.4 (SAS Institute Cambridge, UK). Inc., Cary, NC, USA). Plots were created using R v4.0.2 (R Foundation for Statistical Computing, Vienna, Aus- Blood sample collection and blood serum analysis tria). Pen was considered as the experimental unit for Blood samples were collected via venepuncture of the all performance, serum and faecal data analyses. Alpha external jugular vein (approximately 10 ml/pig) from level for determination of significance was 0.05 and 2 pigs/pen selected randomly at 10 days of the trial trends were identified as alpha of 0.10. Data were tested in S1 (20 weeks of age) and at 14 days of the trial in S2 for normality using the Shapiro–Wilk test and by exam- (14 weeks of age). A total of 40 and 56 (n = 8 per treat- ining the normal probability plot. Initial BW data were ment) blood samples were collected in S1 and S2 respec- analysed using general linear models including treat- tively. Blood samples were collected early in the morning ment as fixed effect. Models for BW, ADG, ADFI, and in a non-fasting state as per commercial practice. Blood FCR variables were analysed using general linear models samples were kept immediately on ice at 4 °C after col- including treatment diet as fixed effect and initial BW lection until serum was separated by centrifugation for as a co-variable. For blood serum, models for albumin, 15 min at 2000 rcf. Blood serum samples were analysed glucose, triglycerides, cholesterol, SUN, creatinine, and the same day using the ABX Pentra 400 Clinical Chem- total protein were analysed using general linear models istry benchtop analyser (HORIBA Medical, Irvine, Cali- including treatment diet as fixed effect. For faecal VFA, fornia, USA) and ABX Pentra 400 re-agents (HORIBA models for acetic, propionic, butyric, valeric, BCFA, and ABX SAS, Montpellier, France) at the Teagasc Chemistry VFA were analysed using general linear models including Lab in Fermoy, Co. Cork, Ireland. Selected blood serum treatment diet as fixed effect. Multiple means compari- metabolites (and techniques) were: albumin (bromo- sons were done using Tukey–Kramer’s correction in all cresol green dye-binding procedure), glucose (hexokinase cases. Results for fixed effects are reported as least square method), triglycerides (enzymatic method), cholesterol means ± standard error mean. (enzymatic photometric test), SUN (enzymatic UV test), Receiver Operating Characteristic (ROC) Curves anal- total protein concentration (Biuret reaction), and creati- ysis was used to identify biomarkers that discriminated nine (enzymatic method). between diets. Data were analysed using the pROC pack- age [34] for R v4.0.2. Univariable ROC curves were calcu- Faecal samples were collected using BioFreeze™ vials (Ali- Faecal sample collection and volatile fatty acid analysis lated for SUN and BCFA comparing LP1 versus HP1, LE1 versus HE1, LP2 versus HP2, LE2 versus HE2, and LA2 metrics Diagnostics Ltd, Espoo, Finland) from 2 pigs/pen versus HA2. The accuracy of the models was assessed selected randomly at 10 days of the trial in S1 (20 weeks by calculating the area under the curve (AUC). Values of age) and at 14 days of the trial in S2 (14 weeks of age). of AUC were interpreted as non-accurate (AUC = 0.5), Faecal samples were collected from pigs which defecated less accurate (0.5 < AUC ≤ 0.7), moderately accurate at the same moment of the sampling. A total of 40 and 56 (0.7 < AUC ≤ 0.9), highly accurate (0.9 < AUC < 1) and Camp Montoro et al. Porcine Health Management (2022) 8:32 Page 6 of 14 perfect (AUC = 1) [35]. The AUC is significant when the Blood serum metabolites confidence interval does not include 50%. Cut-off con- Blood metabolite profile for S1 and S2 are presented in centrations were calculated for each ROC curve and the Tables 5 and 6, respectively. In S1, albumin and glucose corresponding sensitivities and specificities, and 95% did not differ between dietary treatments, although HE1 confidence intervals were obtained [36]. pigs tended to have higher glucose concentration levels than LE1 pigs (P = 0.063). For creatinine and triglycer- ides, HE1 pigs had higher levels than LE1 pigs (P < 0.05) Results however, HE1 pigs had lower total protein levels than C1 Body weight, feed intake, and feed efficiency traits pigs (P = 0.033). Furthermore, LP1 pigs had higher cho- In S1, there were no differences on BW, ADG, ADFI, and lesterol concentration levels than LE1 pigs (P = 0.015) FCR between dietary treatments (Table 3). In S2, pigs while HP1 pigs had higher SUN (13.63 ± 0.951 mg/dL) fed the LA2 diet were 3.5 kg lighter and gained 171.0 g/d than the other dietary treatments (7.47 ± 0.951 mg/dL; less in average than those pigs fed with the C2, LP2, P < 0.001; Fig. 1). In S2, glucose and creatinine did not dif- HP2, HA2, and HE2 diets at the end of the trial (P < 0.01; fer between dietary treatments. For albumin, LA2 pigs Table 4). Moreover, pigs fed the LP2 diet consumed had lower concentrations than pigs fed with HP2, HA2, 229.3 g/d more than HP2 pigs (P = 0.035) at the end of LE2, and HE2 diets (P < 0.001). Also, LA2 pigs had lower the trial. Finally, pigs fed the LA2 diet had higher FCR total protein levels than HP2 pigs (P = 0.05). Pigs fed HP2 than pigs fed with the C2, HP2, HA2, LE2, and HE2 diets diet had higher SUN (11.60 ± 0.613 mg/dL) than the rest (P < 0.001); while HP2 pigs had lower FCR than those pigs of the dietary treatments (7.65 ± 0.613 mg/dL; P < 0.001; fed with the C2, LP2, LA2, and LE2 dietary treatments Fig. 1). while LP2 pigs had lower SUN (5.2 ± 0.61 mg/dL) (P < 0.001) at the end of the trial. than those pigs fed with C2 (8.2 ± 0.61), HA2 (8.7 ± 0.61), Table 3 Productive performance of finishing pigs grouped by dietary treatment in study 1­ 1 Dietary ­treatment2 SEM p-Value Traits C1 LP1 HP1 LE1 HE1 BW 128 days, kg 87.1 88.1 87.8 85.4 86.9 2.32 0.926 BW 138 days, kg 100.2 99.9 101.0 100.2 100.2 0.70 0.828 ADG, g 1315.0 1283.7 1399.1 1312.1 1317.7 71.49 0.825 ADFI, g 3024.2 3078.4 3110.0 3284.8 2977.7 128.86 0.531 FCR 2.33 2.39 2.23 2.48 2.27 0.08 0.228 1 Body weight (BW), average daily gain (ADG), average daily feed intake (ADFI), feed conversion ratio (FCR) (SEM = Standard error mean [SEM]). Pigs were followed from 128 to 138 days of age (n = 4) 2 Dietary treatments: C1 (Control; 10.03 MJ/kg of NE; 160.0 g of CP; 9.5 g of SID Lys per kg of feed), LP1 (Low Protein; 10.03 MJ/kg of NE; 132.0 g of CP; 7.5 g of SID Lys per kg of feed), HP1 (High Protein; 10.03 MJ/kg of NE; 188.0 g of CP; 11.5 g of SID Lys per kg of feed), LE1 (Low Energy; 9.61 MJ/kg of NE; 160.0 g of CP; 9.5 g of SID Lys per kg of feed), and HE1 (High Energy; 10.45 MJ/kg of NE; 160.0 g of CP; 9.5 g of SID Lys per kg of feed) Table 4 Productive performance of growing pigs grouped by dietary treatment in study 2­ 1 Traits Dietary ­treatment2 SEM p-Value C2 LP2 HP2 LA2 HA2 LE2 HE2 BW 84 days, kg 41.1 41.7 41.3 41.5 41.1 41.2 41.3 1.49 1.000 BW 104 days, kg 62.6 a 62.1 a 62.7 a 59.3 b 63.2 a 61.5 ab 63.2 a 0.57 0.001 ADG, g 1061.0 a 1038.9 a 1075.5 a 900.9 b 1090.1 a 1011.5 ab 1093.8 a 28.90 0.002 ADFI, g 1932.3 ab 1950.2 a 1720.9 b 1819.7 ab 1933.2 ab 1850.3 ab 1878.4 ab 46.76 0.030 FCR 1.84 b 1.89 ab 1.61 c 2.03 a 1.78 bc 1.85 b 1.73 bc 0.04 < 0.001 1 Body weight (BW), average daily gain (ADG), average daily feed intake (ADFI) and feed conversion ratio (FCR) (SEM = Standard error mean [SEM]). Pigs were followed from 84 to 104 days of age (n = 4) 2 Dietary treatments: C2 (Control; 10.03 MJ/kg NE; 165.0 g of CP; 9.5 g of SID Lys per kg of feed), LP2 (Low Protein; 10.07 MJ/kg NE; 140.0 g of CP; 9.5 g of SID Lys per kg of feed), HP2 (High Protein; 10.03 MJ/kg NE; 190.0 g of CP; 9.5 g of SID Lys per kg of feed), LA2 (Low Amino Acid; 10.03 MJ/kg NE; 135.0 g of CP; 7.5 g of SID Lys per kg of feed), HA2 (High Amino Acid; 10.07 MJ/kg NE; 165.0 g of CP; 11.5 g of SID Lys per kg of feed), LE2 (Low Energy; 9.61 MJ/kg NE; 165.0 g of CP; 9.5 g of SID Lys per kg of feed), and HE2 (High Energy; 10.45 MJ/kg NE; 165.0 g of CP; 9.5 g of SID Lys per kg of feed) a −b Within rows, significant differences between groups (P < 0.05) Camp Montoro et al. Porcine Health Management (2022) 8:32 Page 7 of 14 Table 5 Blood serum and faecal volatile fatty acid (VFA) profile from 40 finishing pigs grouped by dietary treatment (n = 4) in study 1­ 1 Traits Dietary treatment 2 SEM p-Value C1 LP1 HP1 LE1 HE1 Blood serum Albumin, g/L 36.54 37.08 38.04 36.04 36.53 0.880 0.565 Glucose, mmol/L 5.31 5.24 5.39 4.61 5.68 0.273 0.105 Triglycerides, mmol/L 0.38 ab 0.35 ab 0.37 ab 0.28 b 0.42 a 0.030 0.023 Cholesterol, mmol/L 2.12 ab 2.44 a 2.27 ab 2.03 b 2.24 ab 0.084 0.021 ab ab ab b Creatinine, µmol/L 123.03 127.50 123.85 117.20 133.81 a 3.966 0.070 Total protein, g/L 67.36 a 64.20 ab 63.61 ab 61.46 ab 60.86 b 1.509 0.035 Faecal VFA, % of total VFA Acetic 57.4 b 61.6 a 62.0 a 60.6 ab 58.5 ab 1.01 0.012 Propionic 21.9 21.4 21.2 22.0 21.7 0.59 0.827 Butyric 11.4 9.8 9.3 10.9 11.0 0.55 0.061 Valeric 2.9 a 2.3 ab 2.4 ab 2.1 b 2.6 ab 0.15 0.010 VFA, mmol/kg 193.0 158.2 166.9 170.5 154.7 14.98 0.451 1 Blood and faecal samples were collected from 2 pigs/pen selected randomly at the end of the trial at 20 weeks of age (Means ± Standard error mean [SEM]) 2 Dietary treatments: C1 (Control; 10.03 MJ/kg of NE; 160.0 g of CP; 9.5 g of SID Lys per kg of feed), LP1 (Low Protein; 10.03 MJ/kg of NE; 132.0 g of CP; 7.5 g of SID Lys per kg of feed), HP1 (High Protein; 10.03 MJ/kg of NE; 188.0 g of CP; 11.5 g of SID Lys per kg of feed), LE1 (Low Energy; 9.61 MJ/kg of NE; 160.0 g of CP; 9.5 g of SID Lys per kg of feed), and HE1 (High Energy; 10.45 MJ/kg of NE; 160.0 g of CP; 9.5 g of SID Lys per kg of feed) a −b Within rows, significant differences between groups (P < 0.05) Table 6 Blood serum and faecal volatile fatty acid (VFA) profile from 56 growing pigs grouped by dietary treatment (n = 4) in study 2­ 1 Traits Dietary treatment 2 SEM p-Value C2 LP2 HP2 LA2 HA2 LE2 HE2 Blood serum Albumin, g/L 36.04ab 34.76ab 38.05a 32.10b 39.08a 37.25a 37.19a 1.083 < 0.001 Glucose, mmol/L 5.37 5.83 5.72 5.94 6.05 5.69 5.57 0.253 0.559 Triglycerides, mmol/L 0.33b 0.55ab 0.44ab 0.57ab 0.47ab 0.67a 0.61a 0.059 0.004 b ab ab ab ab b Cholesterol, mmol/L 2.34 2.61 2.45 2.52 2.39 2.35 2.80a 0.107 0.046 Creatinine, µmol/L 111.01 100.29 117.56 113.89 112.97 113.00 107.34 4.422 0.168 Total protein, g/L 61.98ab 64.60ab 65.95a 59.98b 63.49ab 61.64ab 63.44ab 1.391 0.077 Faecal VFA, % of total VFA Acetic 55.99 55.22 54.64 54.66 56.84 56.37 56.01 1.192 0.798 Propionic 21.15b 23.18ab 21.40b 24.53a 21.19b 21.57b 21.51b 0.567 < 0.001 Butyric 13.53 12.57 13.25 11.81 13.42 13.04 13.42 0.710 0.608 Valeric 4.08 3.96 4.09 3.99 3.72 4.18 3.49 0.367 0.844 VFA, mmol/kg 93.78 88.44 91.55 86.58 95.75 99.86 88.38 6.281 0.754 1 Blood and faecal samples were collected from 2 pigs/pen selected randomly at day 14 of the trial at 14 weeks of age (Means ± Standard error mean [SEM]) 2 Dietary treatments: C2 (Control; 10.03 MJ/kg NE; 165.0 g of CP; 9.5 g of SID Lys per kg of feed), LP2 (Low protein; 10.07 MJ/kg NE; 140.0 g of CP; 9.5 g of SID Lys per kg of feed), HP2 (High protein; 10.03 MJ/kg NE; 190.0 g of CP; 9.5 g of SID Lys per kg of feed), LA2 (Low amino acid; 10.03 MJ/kg NE; 135.0 g of CP; 7.5 g of SID Lys per kg of feed), HA2 (High amino acid; 10.07 MJ/kg NE; 165.0 g of CP; 11.5 g of SID Lys per kg of feed), LE2 (Low energy; 9.61 MJ/kg NE; 165.0 g of CP; 9.5 g of SID Lys per kg of feed), and HE2 (High energy; 10.45 MJ/kg NE; 165.0 g of CP; 9.5 g of SID Lys per kg of feed) a −b Within rows, significant differences between groups (P < 0.05) LE2 (8.8 ± 0.61), HE2 (8.5 ± 0.61), and HP2 (P < 0.001). Volatile fatty acids profile Finally, C2 pigs showed lower triglycerides levels than Total VFA and VFA profiles for S1 and S2 are presented pigs fed with LE2 and HE2 dietary treatments (P < 0.01); in Tables 5 and 6, respectively. In S1, total VFA (mmol/ however, HE2 pigs had higher cholesterol concentration kg) did not differ between dietary treatments. Pigs fed levels than C2 and LE2 pigs (P < 0.05). C1 had lower percentage of acetic than LP1 and HP1 Camp Montoro et al. Porcine Health Management (2022) 8:32 Page 8 of 14 Fig. 1 Serum urea nitrogen (SUN; mg/dL) and branched-chain fatty acids (BCFA; % of total VFA) levels grouped by dietary treatment (n = 4) from study 1 (A, C) and study 2 (B, D). Results are presented as Means ± Standard error mean. a, b Significant differences between treatments (P < 0.05). Blood and faecal samples were collected from 2 pigs/pen selected randomly at 20 weeks of age (study 1) and 14 weeks of age (study 2). Dietary treatments from study 1: C1 (Control; 10.03 MJ/kg of NE; 160.0 g of CP; 9.5 g of SID Lys per kg of feed), LP1 (Low Protein; 10.03 MJ/kg of NE; 132.0 g of CP; 7.5 g of SID Lys per kg of feed), HP1 (High Protein; 10.03 MJ/kg of NE; 188.0 g of CP; 11.5 g of SID Lys per kg of feed), LE1 (Low Energy; 9.61 MJ/ kg of NE; 160.0 g of CP; 9.5 g of SID Lys per kg of feed), and HE1 (High Energy; 10.45 MJ/kg of NE; 160.0 g of CP; 9.5 g of SID Lys per kg of feed). Dietary treatments from study 2: C2 (Control; 10.03 MJ/kg NE; 165.0 g of CP; 9.5 g of SID Lys per kg of feed), LP2 (Low protein; 10.07 MJ/kg NE; 140.0 g of CP; 9.5 g of SID Lys per kg of feed), HP2 (High protein; 10.03 MJ/kg NE; 190.0 g of CP; 9.5 g of SID Lys per kg of feed), LA2 (Low amino acid; 10.03 MJ/kg NE; 135.0 g of CP; 7.5 g of SID Lys per kg of feed), HA2 (High amino acid; 10.07 MJ/kg NE; 165.0 g of CP; 11.5 g of SID Lys per kg of feed), LE2 (Low energy; 9.61 MJ/kg NE; 165.0 g of CP; 9.5 g of SID Lys per kg of feed), and HE2 (High energy; 10.45 MJ/kg NE; 165.0 g of CP; 9.5 g of SID Lys per kg of feed) pigs (P < 0.05) but higher percentage of valeric than LE1 ROC curve analysis pigs (P = 0.010). Pigs fed LE1 showed lower percentage The ROC curve analysis for S1 is presented in Fig. 2. of BCFA (4.4 ± 0.38) of total VFA than C1 and HE1 pigs The AUC for SUN was of 98.4% with two optimal cut- (6.4 ± 0.38; 6.1 ± 0.38, respectively; P < 0.01; Fig. 1). In offs of 9.4 mg/dL (100% sensitivity, 87.5% specificity) S2, total VFA (mmol/kg) did not differ between dietary and 11.2 mg/dL (87.5% sensitivity, 100% specificity) that treatments (P > 0.05). Moreover, acetic, butyric, and could serve to differentiate the LP1 versus HP1 dietary valeric (as % of total VFA) did not differ between die- treatments. Branched-chain fatty acids showed a high tary treatments (P > 0.05). Nevertheless, LA2 pigs had accuracy (AUC = 96.9%) to differentiate the LE1 versus higher percentage of propionic of total VFA than those HE1 dietary treatments. The ROC curve analysis for S2 pigs fed with C2, HP2, HA2, LE2, and HE2 dietary is presented in Fig. 3. The AUC for SUN was 100% with treatments (P < 0.001). Finally, HP2 pigs had higher per- an optimal cut-off of 9.4 mg/dL to differentiate LP2 ver- centage of BCFA (6.61 ± 0.408) of total VFA than HA2 sus HP2 dietary treatments. Branched-chain fatty acids and LE2 pigs (4.84 ± 0.408; 4.84 ± 0.408, respectively; showed a moderate-high accuracy to differentiate LP2 P < 0.05; Fig. 1); and HP2 pigs tended to have higher versus HP2, and LE2 versus HE2 dietary treatments percentage of BCFA of total VFA than LP2 and LA2 (AUC = 87.5% and 81.2%, respectively). pigs (5.06 ± 0.408; 5.01 ± 0.408, respectively; P < 0.10; Fig. 1). Camp Montoro et al. Porcine Health Management (2022) 8:32 Page 9 of 14 Fig. 2 Receiver operating characteristic (ROC) curves for serum urea nitrogen (SUN) and branched-chain fatty acids (BCFA) biomarkers used to differentiate low versus high protein and energy diets in finishing pigs at 20 weeks of age (Study 1). Dietary treatments from study 1: low Protein (10.03 MJ/kg of NE; 132.0 g of CP; 7.5 g of SID Lys per kg of feed), High Protein (10.03 MJ/kg of NE; 188.0 g of CP; 11.5 g of SID Lys per kg of feed), Low Energy (9.61 MJ/kg of NE; 160.0 g of CP; 9.5 g of SID Lys per kg of feed), and High Energy (10.45 MJ/kg of NE; 160.0 g of CP; 9.5 g of SID Lys per kg of feed). Headings include the area under the curve (AUC) and the 95% confidence interval (CI). The AUC is significant when the CI does not include 50%. Within each graph, the optimal cut-off concentration and the corresponding specificity and sensitivity (parenthesis) are shown (See figure on next page.) Fig. 3 Receiver operating characteristic (ROC) curves for serum urea nitrogen (SUN) and branched-chain fatty acids (BCFA) biomarkers used to differentiate low versus high protein, amino acids, and energy diets in growing pigs at 14 weeks of age (Study 2). Dietary treatments from study 2: Low protein (10.07 MJ/kg NE; 140.0 g of CP; 9.5 g of SID Lys per kg of feed), High protein (10.03 MJ/kg NE; 190.0 g of CP; 9.5 g of SID Lys per kg of feed), Low amino acid (10.03 MJ/kg NE; 135.0 g of CP; 7.5 g of SID Lys per kg of feed), High amino acid (10.07 MJ/kg NE; 165.0 g of CP; 11.5 g of SID Lys per kg of feed), Low energy (9.61 MJ/kg NE; 165.0 g of CP; 9.5 g of SID Lys per kg of feed), and High energy (10.45 MJ/kg NE; 165.0 g of CP; 9.5 g of SID Lys per kg of feed). Headings include the area under the curve (AUC) and the 95% confidence interval (CI). The AUC is significant when the CI does not include 50%. Within each graph, the optimal cut-off concentration and the corresponding specificity and sensitivity (parenthesis) are shown Camp Montoro et al. Porcine Health Management (2022) 8:32 Page 10 of 14 Fig. 3 (See legend on previous page.) Camp Montoro et al. Porcine Health Management (2022) 8:32 Page 11 of 14 Discussion AAs. The use of SUN is also an advantage because of its Although these trials were not designed to study differ- short time to achieve a constant concentration in blood ences in productive performance due to the short period after changing the diet [20]. Further research is needed of study and low sample size, productive performance to fully understand the applicability of SUN in a cohort of was monitored and there were some interesting findings. commercial farms. Since there is no control diets at farm In S1, there were no differences in productive perfor- level, an empirical approach should be conducted within mance among diets. Thus, for S2, the authors decided to a farm and between farms to establish SUN standard use younger pigs which are more likely to be affected by intervals based on regressions and considering factors reductions in dietary AA levels similar to those used in S1 such as diets, feed form, age or BW, among others, when (from 1.15 to 0.75 g SID lysine per MJ). Pigs fed LA2 diet relevant. The infrastructure required for such analysis were the lightest and had the worst efficiency of all the would be similar to those offered currently for small ani- diets in S2, probably because the levels of AAs were not mal medicine where SUN and other biomarkers are used enough to meet the requirements of the pigs [26]. Pigs in regularly. S2 were also followed for a slightly longer period of time Total protein and albumin are also involved in protein which may have allowed them to show differences in pro- metabolism and were studied as interesting biomarkers. ductive performance. It is interesting to note that pro- There were no differences in S1 in total protein and albu- ductive performance of pigs fed LP2 diet was not affected min between diets. The latter is in agreement with Regmi compared to HP2. This diet was supplemented with AAs et al. [17], who did not observe differences in serum total and both LP2 and HP2 had the same level of SID lysine protein concentration in finishing pigs, of similar age to (9.5 g per MJ). Thus, LP2 would achieve a lower risk of those in S1, fed insufficient (0.32%), adequate (0.60%) or abnormal protein fermentations and lower emissions excessive (0.87%) SID lysine diets during 4 weeks, and than HP2 with no negative effects in performance. only observed reduced plasma albumin concentration in Serum metabolites have been used in pigs for research finishing pigs when fed the 0.32% SID lysine diet which is purposes but the literature on their use for clinical pur- far below the SID lysine levels of the dietary treatments poses is scarce and reference values are needed consid- of the present study. Nevertheless, in S2, pigs fed higher ering different factors such as age, breed, sex, diet, and amounts of CP showed an increase in serum total protein methods of sample collection and analysis [37]. In this and albumin levels. These findings are in accordance with study, the authors aimed to use serum metabolites as bio- some of the previous literature [16, 18]. Thus, the age of markers to discriminate between diets differing on CP, the pig may affect serum total protein and albumin lev- AA, and NE levels with the final intention to use these els and finishing pigs may be able to show a homeostatic biomarkers in daily practice as indicator of suboptimal control besides the dietary CP content. Moreover, finish- diets. In S1, SUN was the clearest indicator of differ- ing pigs have already reached the maximum protein dep- ences between diets with high and low CP levels, which osition [39–41] and their metabolism may not be focused were discriminated in the ROC curve analysis with an on protein turn-over, contrary to early stages of the AUC close to 1. The principal end product of protein grower-finisher period. Albumin was a good biomarker catabolism is SUN [38], thus it makes sense that SUN to differentiate between LA2 and HA2 in S2, but it was increases when the diet has an excess of CP that cannot not as consistent as SUN. Further research is needed to be used by the animal due to CP excess or AA imbal- explore its use in multivariable biomarkers in combina- ances. Diets LP1 and HP1 differed in both factors, the tion with SUN. level of CP and the level of AAs. To separate the effect Serum creatinine is also related to protein metabolism of these 2 factors, in S2, LP2 and HP2 were formulated because is the product from muscle metabolism [37] and to have different levels of CP but the same levels of AAs, has a positive correlation with total and striated mus- and LA2 and HA2 were formulated to differ on AA lev- cle [42]. In the current studies, creatinine did not show els without an excess on protein. As expected, diet HP2, any clear patterns and may not be as good as SUN as a with an excess of CP, resulted in an increased SUN level biomarker. compared to LP2. However, diet HA2 compared to LA2 Concerning energy, serum glucose did not show did not induce the same increase in SUN that HP2 com- clear patterns between dietary treatments either which pared to LP2, despite having a similar increase in CP. The agrees with previous literature [15–17] and shows a increase in CP in diet HA2 could be used by the pig for good homoeostatic control of serum glucose concentra- growth because it was achieved by increasing AAs levels tion by either growing and finishing pigs in commercial according to ideal protein profile. Overall, SUN may be a conditions fed ad libitum. Triglycerides and choles- useful indicator of protein efficiency at farm level when terol are both metabolites involved in lipid metabolism pigs are fed suboptimal dietary CP diets not balanced in [37]. They did not show any consistent pattern despite Camp Montoro et al. Porcine Health Management (2022) 8:32 Page 12 of 14 showing some differences between diets differening Conclusion in NE levels. The inconsistency in the results may be This study aimed to assess the feasibility of blood serum related to the age of the animals. The hypercholester- metabolite and faecal VFA profiles as potential biomark- olemic effect observed in S2 pigs but not so clear in ers to identify differences in the dietary levels of energy, S1 is in agreement with previous literature [16–18], protein, and amino acids in growing and finishing pigs at although the exact mecanishim of this effect is not clear farm level when fed suboptimal diets. Out of all the blood yet. Further research should be carried out in order to serum metabolites studied, SUN seems to be the best further define these results. indicator to assess protein efficiency at farm level show- No differences in total VFAs concentrations were ing high accuracy to detect an excess of crude protein observed between the dietary treatments in any of the with an inadequate amino acid profile in growing and fin- two trials, and none of the individual VFAs showed dif- ishing pig diets. Regarding the faecal VFA profile, BCFAs ference worth discussing except for BCFA. The authors could be a potential indicator for high CP diets but may hypothesised that BCFA would show CP excess in the be affected by the age of the pig. Further studies at com- diets based on the previous literature [23, 25]. However, mercial scale are needed to fully understand the applica- the pattern differed between trials. Growing pigs in S2 bility of SUN in a cohort of commercial farms. The use fed HP2 diet showed a higher percentage of BCFA than of other biomarkers as part of a multivariable indicator pigs fed the HA2 and LE2 diets. Moreover, HP2 pigs should also be considered. had numerically greatly amounts of BCFA compared to those pigs fed the LP2 and LA2 diets. ROC curve analy- Abbreviations sis showed that BCFA has a moderate-high accuracy to AA: Amino Acids; ADFI: Average Daily Feed Intake; ADG: Average Daily Gain; differentiate LP2 and HP2 diets in growing pigs. These BCFA: Branched-Chain Fatty Acids; BW: Body Weight; C1: Control Study 1; C2: findings agree with previous literature that reported an Control Study 2; CP: Crude Protein; FCR: Feed Conversion Ratio; HA2: High Amino Acids Study 2; HE1: High Net Energy Study 1; HE2: High Net Energy increased production of BCFA in manure in grower-fin- Study 2; HP1: High Crude Protein Study 1; HP2: High Crude Protein Study 2; isher pigs fed high CP diets [23, 43]. The low production LA2: Low Amino Acids Study 2; LE1: Low Net Energy Study 1; LE2: Low Net of BCFA in pigs fed the HA2 diet compared to those fed Energy Study 2; LP1: Low Crude Protein Study 1; LP2: Low Crude Protein Study 2; Lys: Lysine; NE: Net Energy; S1: Study 1; S2: Study 2; SID: Standardized Ileal the HP2 diet might be related to a fast absorption rate of Digestible; SUN: Serum Urea Nitrogen; VFA: Volatile Fatty Acids. free AA and a lower level of CP available for fermenta- tion which agrees with the pattern observed for SUN. In Acknowledgements The authors gratefully acknowledge the Teagasc Pig Research Facility staff for S1, finishing pigs fed the HP1 diet did not show higher their help on this study. percentage of BCFA than any other dietary treatment. This absence of differences in BCFA might be explained Author contributions J.C.M and E.G.M conceived and designed the study. J.C.M collected the data. by the fact that older pigs have a more developed gas- J.C.M and E.G.M analysed the data. J.C.M drafted the manuscript. E.G.M, D.S.O, trointestinal tract with a high fermentation capacity that R.M, J.G and N.L. reviewed and edited the manuscript. E.G.M supervised makes it difficult to observe differences between the die- the study. All authors have read and agreed to the published version of the manuscript. tary treatments at faecal level. The differences may exist in cecum or proximal colon but are not present in faeces. Funding On the other hand, a higher percentage of BCFA between Jordi Camp Montoro was supported by the Teagasc Walsh Scholarship Scheme grant-in-aid project Introducing new feeding programs and facilities for pigs was found in pigs fed diet HE1 when compared to Irish finishing pigs ref 0415. The funders had no role in the study design, data LE1. This difference could be related to the added fibre collection and analysis, decision to publish, or preparation of the manuscript. in the LE1 diet. In this line, a recent study reported that Availability of data and materials the increased body weight and age of the pigs resulted in The datasets used and analysed during the current study are available from an improved digestibility of dietary fibre fractions [44], the corresponding author on reasonable request. which will influence the VFA profile as it is positively correlated with the apparent total tract digestibility of Declarations insoluble dietary fibre and cellulose [45]. Therefore, the Ethics approval and consent to participate fermentation of soybean hulls could have produced a The study received ethical approval from the Teagasc Animal Ethics Com‑ shift in the VFA profile reducing the BCFA production mittee (TAEC 244/2019). The authors confirm that they have followed the by the microbial population. Although BCFA did not European Union Directive 2010/63 on the protection of animals used for scientific purposes. show the same consistency as SUN, more research is warranted. The deamination of branched AAs may also Consent for publication cause a shift in the microbiome population to increase Not applicable. production of BCFA [25] which may be a more sensitive Competing interests biomarker. The authors declare that they have no competing interests. Camp Montoro et al. Porcine Health Management (2022) 8:32 Page 13 of 14 Author details 18. Mule HR, Chiba LI, Fabian J, Kuhlers DL, Jungst SB, Frobish LT, et al. Effect 1 Pig Development Department, Animal and Grassland Research and Innova‑ of early dietary amino acid restrictions on serum metabolites in pigs tion Centre, Teagasc, Moorepark, Fermoy, Co. Cork P61 C996, Ireland. 2 Depart‑ selected for lean growth efficiency. Can J Anim Sci. 2006;86(4):489–500. ment of Animal and Food Sciences, Animal Nutrition and Welfare Service, 19. Chen HY, Lewis AJ, Miller PS, Yen JT. The effect of excess protein on Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain. 3 Agri-Food growth performance and protein metabolism of finishing barrows and and Biosciences Institute, Large Park, HillsboroughBT 26 6DR, Co Down, North‑ gilts. J Anim Sci. 1999;77(12):3238–47. ern Ireland, UK. 4 Cooperativa d’Ivars d’Urgell SCCL, 25260 Ivars d’Urgell, Lleida, 20. Coma J, Carrion D, Zimmerman DR. Use of plasma urea nitrogen as a Spain. 5 UCD Veterinary Sciences Centre, University College Dublin, Belfield, rapid response criterion to determine the lysine requirement of pigs. J Dublin 4 D04 V1W8, Ireland. Anim Sci. 1995;73(2):472–81. 21. Hong JS, Lee GI, Jin XH, Kim YY. Effect of dietary energy levels and phase Received: 23 February 2022 Accepted: 20 June 2022 feeding by protein levels on growth performance, blood profiles and carcass characteristics in growing-finishing pigs. J Anim Sci Technol. 2016;58(1):1–10. 22. Kerr BJ, Kellner TA, Shurson GC. Characteristics of lipids and their feeding value in swine diets. J Anim Sci Biotechnol. 2015;6(1):1–23. References 23. Cho S, Hwang O, Park S. Effect of dietary protein levels on composition of 1. Shurson GC, Hung YT, Jang JC, Urriola PE. Measures matter—determining odorous compounds and bacterial ecology in pig manure. Asian-Austral‑ the true nutri-physiological value of feed ingredients for swine. Animals. asian J Anim Sci. 2015;28(9):1362–70. 2021;11(5):1–50. 24. Dahl WJ, Rivero Mendoza D, Lambert JM. Diet, nutrients and the microbi‑ 2. O’Meara FM, Gardiner GE, O’Doherty JV, Lawlor PG. The effect of feed ome, Progress in molecular biology and translational science, 1st edn, vol. form and delivery method on feed microbiology and growth perfor‑ 171. Amsterdam: Elsevier; 2020. p. 237–63. mance in grow-finisher pigs. J Anim Sci. 2020;98(3):1–11. 25. Le PD, Aarnink AJA, Ogink NWM, Becker PM, Verstegen MWA. Odour 3. Camp Montoro J, Solà-Oriol D, Muns R, Gasa J, Llanes N, Manzanilla EG. from animal production facilities: its relationship to diet. Nutr Res Rev. High levels of standardized ileal digestible amino acids improve feed effi‑ 2005;18(1):3–30. ciency in slow-growing pigs at late grower-finisher stage. J Anim Physiol 26. NRC. Nutrient requirements of swine. Washington, DC: National Acad‑ Anim Nutr (Berl). 2021;00(February):1–8. emies Press; 2012. 4. Camp Montoro J, Manzanilla EG, Solà-Oriol D, Muns R, Gasa J, Clear O, 27. van Milgen J, Dourmad JY. Concept and application of ideal protein for et al. Predicting productive performance in grow-finisher pigs using birth pigs. J Anim Sci Biotechnol. 2015;6(1):1–11. and weaning body weight. Animals. 2020;10(6):1–14. 28. Thiex N. Evaluation of analytical methods for the determination of mois‑ 5. Camp Montoro J, Pessoa J, Sol D, Muns R, Gasa J, Manzanilla EG. Effect of ture, crude protein, crude fat, and crude fiber in distillers dried grains with phase feeding, space allowance and mixing on productive performance solubles. J AOAC Int. 2009;92(1):61–73. of grower-finisher pigs. Animals. 2022;12(3):1–14. 29. Thiex N, Novotny L, Crawford A. Determination of ash in animal feed: 6. Camp Montoro J, Boyle LA, Solà-Oriol D, Muns R, Gasa J, Garcia ME. Effect AOAC official method 942.05 revisited. J AOAC Int. 2012;95(5):1392–7. of space allowance and mixing on growth performance and body lesions 30. Ebeling ME. The dumas method for nitrogen in feeds. J AOAC Int. of grower-finisher pigs in pens with a single wet-dry feeder. Porc Heal 1968;51(4):766–70. Manag. 2021;7(7):1–8. 31. Thiex NJ, Anderson S, Gildemeister B. Crude fat, diethyl ether extraction, 7. van der Meer Y, Lammers A, Jansman AJM, Rijnen MMJA, Hendriks WH, in feed, cereal grain, and forage (Randall/Soxtec/Submersion method): Gerrits WJJ. Performance of pigs kept under different sanitary conditions collaborative study. J AOAC Int. 2003;86(5):888–98. affected by protein intake and amino acid supplementation. J Anim Sci. 32. Otter DE. Standardised methods for amino acid analysis of food. Br J Nutr. 2016;94(11):4704–19. 2012;108(SUPPL. 2):230–7. 8. Van Der Meer Y, Jansman AJM, Gerrits WJJ. Low sanitary conditions 33. Czerkawski JW. The use of pivalic acid as a reference substance in meas‑ increase energy expenditure for maintenance and decrease incremental urements of production of volatile fatty acids by rumen micro-organisms protein efficiency in growing pigs. Animal. 2020;14(9):1811–920. in vitro. Br J Nutr. 1976;36(2):311–5. 9. Rist VTS, Weiss E, Eklund M, Mosenthin R. Impact of dietary protein on 34. Turck N, Vutskits L, Sanchez-Pena P, Robin X, Hainard A, Gex-Fabry M, et al. microbiota composition and activity in the gastrointestinal tract of pROC: an open-source package for R and S+ to analyze and compare piglets in relation to gut health: a review. Animal. 2013;7(7):1067–78. ROC curves. BMC Bioinform. 2011;8:12–77. 10. Prandini A, Sigolo S, Morlacchini M, Grilli E, Fiorentini L. Microencap‑ 35. Greiner M, Pfeiffer D, Smith RD. Principles and practical application of the sulated lysine and low-protein diets: effects on performance, carcass receiver-operating characteristic analysis for diagnostic tests. Prev Vet characteristics and nitrogen excretion in heavy growing-finishing pigs. J Med. 2000;45(1–2):23–41. Anim Sci. 2013;91(9):4226–34. 36. Fluss R, Faraggi D, Reiser B. Estimation of the Youden Index and its associ‑ 11. Le Goff G, Noblet J. Comparative total tract digestibility of dietary ated cutoff point. Biometrical J. 2005;47(4):458–72. energy and nutrients in growing pigs and adult sows. J Anim Sci. 37. Constable PD, Hinchcliff KW, Done SH, Grünberg W. Veterinary medicine 2001;79(9):2418–27. a textbook of the diseases of cattle, horses, sheep, pigs and goats, 11th 12. Patience JF, Engele K, Beaulieu AD, Gonyou HW, Zijlstra RT. Variation: costs edn, vol. 1–2. Missouri: Elsevier; 2017. and consequences. Adv Pork Prod. 2004;15(Cv):257–66. 38. Wu G. Amino acids: biochemistry and nutrition. Boca Raton: CRC Press; 13. Patience JF, Rossoni-Serão MC, Gutiérrez NA. A review of feed efficiency 2013. in swine: biology and application. J Anim Sci Biotechnol. 2015;6(1):1–9. 39. Brossard L, Dourmad JY, Rivest J, Van Milgen J. Modelling the variation in 14. Jongbloed AW, Lenis NP. Alteration of nutrition as a means to reduce performance of a population of growing pig as affected by lysine supply environmental pollution by pigs. Livest Prod Sci. 1992;31(1–2):75–94. and feeding strategy. Animal. 2009;3(8):1114–23. 15. Zeng PL, Yan HC, Wang XQ, Zhang CM, Zhu C, Shu G, et al. Effects of 40. Remus A, Hauschild L, Methot S, Pomar C. Precision livestock farming: dietary lysine levels on apparent nutrient digestibility and serum amino real-time estimation of daily protein deposition in growing–finishing acid absorption mode in growing pigs. Asian-Australasian J Anim Sci. pigs. Animal. 2020;14:s360–70. 2013;26(7):1003–11. 41. van Milgen J, Valancogne A, Dubois S, Dourmad JY, Sève B, Noblet J. 16. Kamalakar RB, Chiba LI, Divakala KC, Rodning SP, Welles EG, Bergen InraPorc: a model and decision support tool for the nutrition of growing WG, et al. Effect of the degree and duration of early dietary amino acid pigs. Anim Feed Sci Technol. 2008;143(1–4):387–405. restrictions on subsequent and overall pig performance and physical and 42. Baxmann AC, Ahmed MS, Marques NC, Menon VB, Pereira AB, Kirsztajn sensory characteristics of pork. J Anim Sci. 2009;87(11):3596–606. GM, et al. Influence of muscle mass and physical activity on serum 17. Regmi N, Wang T, Crenshaw MA, Rude BJ, Liao SF. Effects of dietary lysine and urinary creatinine and serum cystatin C. Clin J Am Soc Nephrol. levels on the concentrations of selected nutrient metabolites in blood 2008;3(2):348–54. plasma of late-stage finishing pigs. J Anim Physiol Anim Nutr (Berl). 2018;102(2):403–9. Camp Montoro et al. Porcine Health Management (2022) 8:32 Page 14 of 14 43. Le PD, Aarnink AJA, Jongbloed AW, Van Der Peet-Schwering CMC, Ogink NWM, Verstegen MWA. Effects of dietary crude protein level on odour from pig manure. Animal. 2007;1(5):734–44. 44. Zhao J, Liu X, Zhang Y, Liu L, Wang J, Zhang S. Effects of body weight and fiber sources on fiber digestibility and short chain fatty acid concentra‑ tion in growing pigs. Asian-Australasian J Anim Sci. 2020;33(12):1975–84. 45. Zhao J, Bai Y, Zhang G, Liu L, Lai C. Relationship between dietary fiber fermentation and volatile fatty acids’ concentration in growing pigs. Animals. 2020;10(2):1–14. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub‑ lished maps and institutional affiliations. Ready to submit your research ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions