All procedures involving animals were reviewed and approved by the Ethics Committee for Animal Experiments of the Narmaev Kalmyk Research Institute of Agriculture (Approval No. 2, October 2, 2024). The experimental protocol covered animal selection, housing and management, clinical examination, blood collection, genotyping, growth monitoring, and subsequent slaughter and sample collection for meat-quality assessment. Throughout the study, animal handling and experimental procedures were conducted with consideration for animal welfare and with measures to minimize unnecessary stress and discomfort.

Before enrollment, all bull calves underwent clinical examination to confirm their health status and were vaccinated according to the recommended schedule. Blood samples were collected by trained personnel through jugular venipuncture under standard farm conditions, with approximately 10 mL collected per sampling session. The sampling volume and schedule were selected to meet the analytical requirements of the study while minimizing the burden on the animals. Following blood collection, the animals were monitored for adverse reactions and provided appropriate care when required. No invasive surgical procedures were performed during the experiment.

During the 7-month experimental period, the animals were maintained under standardized husbandry conditions, with access to appropriate feed and drinking water and routine monitoring of their health and welfare. Husbandry and handling procedures were performed by trained personnel to minimize stress. Slaughter of the selected animals was performed at 15 months of age in accordance with the standard meat industry procedure GOST 34120-2017 and using procedures intended to ensure humane treatment. All animals completed the experimental protocol, and no complications affecting animal welfare or study data were reported.

The study was conducted from November 1, 2024, to June 1, 2025, at the evaluation station of the Regional Research and Production Center for Livestock Reproduction, Kalmyk State University named after B.B. Gorodovikov, Elista, Republic of Kalmykia, Russia. The evaluation station is used for comprehensive assessment of breed and production traits in farm animals through breeding and genetic analysis. Throughout the experimental period, the bull calves were housed and managed in accordance with the applicable requirements of the Russian Ministry of Agriculture.

The study included 50 Kalmyk bull calves monitored from 8 to 15 months of age over a 7-month period. The inclusion criteria were established before the study. Eligible animals were Kalmyk bull calves at the post-weaning stage, 8 months of age, and weighing 190–200 kg at enrollment to minimize baseline variability. Until 8 months of age, the calves were reared under the cow-calf production system typical of Kalmyk beef cattle. Only animals without clinical signs of infection, injury, or chronic disease were included. All calves originated from breeding farms in the Republic of Kalmykia and had confirmed pedigrees documented by breeding certificates. Before enrollment, all animals (n = 50) were vaccinated according to the recommended schedule and underwent clinical examination to confirm their health status.

Blood samples were collected by trained personnel through jugular venipuncture under standard farm conditions. Approximately 10 mL of blood was collected per sampling session, with the volume and sampling schedule selected to meet the requirements of the study while minimizing the burden on the animals. After each blood collection, the animals were monitored for adverse reactions and provided appropriate care as needed. No invasive surgical procedures were performed.

Genotyping of the GH c.2141C>G polymorphism was performed using Sanger sequencing. Based on the genotyping results, 37 animals had the LL genotype, and 13 had the LV genotype. Eight bull calves from each genotype were selected for meat-quality assessment, with selection accounting for the mean live weight within each genotype-group to minimize intergroup variation. The sample size of eight animals per genotype was determined a priori based on an anticipated effect size of Cohen's d > 0.8, statistical power of 0.80, and significance level of α_adj = 0.05.

Before the experiment, all 50 Kalmyk bull calves were obtained from five breeding farms in the Republic of Kalmykia. The animals were vaccinated according to the recommended schedule and clinically examined to confirm their health status. Following genotyping, 37 animals were identified as LL and 13 as LV. At the university evaluation station, the animals were housed in two sections, with 25 animals in each section. Allocation to sections was performed using stratified randomization according to the farm of origin; genotype was not considered during allocation. Consequently, animals with different genotypes were co-housed within each section rather than separated into genotype-specific groups. Baseline age and live weight did not differ significantly between the sections (p > 0.05), supporting their comparability.

All animals were maintained under standardized housing and management conditions. Each section had an area of 75 m² and consisted of well-ventilated pens with straw bedding that was cleaned daily. Feed and water were provided ad libitum, and trained personnel performed husbandry and feeding procedures to minimize animal stress. Animals were randomly allocated to pens, and their pen locations remained unchanged between measurement periods. No buffer zones were maintained between groups and contact through pen partitions was possible. Microenvironmental factors, including proximity to feeders and ventilation openings, were not specifically controlled.

Baseline data recorded for each animal included species, sex, age, live weight, farm of origin, and clinical examination findings. Live weight was measured at 8, 12, and 15 months of age, and these measurements were used to calculate average daily gain. Feed and water intake were also recorded. All measurements were performed within a standardized time window from 8:00 to 11:00 AM.

For meat-quality assessment, eight bull calves were selected from each genotype group, for a total sample of 16 animals. Selection accounted for mean live weight within each genotype group to reduce intergroup variability and improve the sensitivity of statistical comparisons. The sample size was determined a priori based on an expected effect size of Cohen's d > 0.8, statistical power of 0.80, and significance level of α_adj = 0.05.

Slaughter was performed at 15 months of age in accordance with standard meat industry procedures (GOST 34120-2017) using methods intended to ensure humane treatment of the animals. Before slaughter, the selected calves (n = 16) underwent a 15-h lairage period, with water withheld during the final 3 h. Following slaughter, meat-quality characteristics were comprehensively evaluated, including morphological assessment of the carcass. Muscle samples intended for chemical analysis were submitted to the V.M. Gorbatov Federal Research Center for Food Systems of the Russian Academy of Sciences according to standardized protocols. All animals completed the experimental protocol, all measurements were included in the analysis, and no complications affecting data quality occurred during the study.

All bull calves received the same diet throughout the experimental period. The diet consisted of mixed-grass hay provided at 4 kg/animal/day and compound feed comprising ground wheat, barley, corn, wheat meal, and a protein-mineral-vitamin complex. Compound feed was supplied using a bunker-type feeding system that provided ad libitum access to the feeders. Drinking water was also available ad libitum throughout the study.

DNA was extracted using an automated procedure with the MagnoPrime VET commercial reagent kit (NextBio LLC, Moscow, Russia). The quality of the extracted DNA was assessed using a Nano-500 spectrophoto-meter. Sanger sequencing was used to determine the GH c.2141C>G genotype.

The sequencing workflow comprised five principal steps. First, polymerase chain reaction (PCR) amplification was performed using the specific oligonucleotide primers 5′-TAGGGGAGGGTGGAAAATGGA-3′ (forward) and 5′-GACACCTACTCAGACAATGCG-3′ (reverse). Second, the PCR products were purified using the ColGen kit (Syntol, Moscow, Russia), and the concentrations of the purified products were determined using a Nano-500 spectrophotometer, ranging from 49.2 to 69.5 ng/µL. Third, Sanger sequencing was performed using the GenSeq Sanger DNA sequencing reagent kit (Syntol). Fourth, the sequencing products were purified using SeqMAG (Syntol). Finally, capillary electrophoresis was performed using a Nanofor 05 genetic analyzer with PDMA-6 sequencing polymer and 1× TAPS sequencing buffer (Syntol).

The PCR reaction mixture had a total volume of 25 µL and was prepared according to the manufacturer's recommendations for Taq DNA polymerase (Evrogen, Moscow, Russia). The mixture contained 2.5 µL of 10× buffer, 1 µL of forward primer (10 µM), 1 µL of reverse primer (10 µM), 1.25 µL of deoxyribonucleotide triphosphate mixture (4 mM), 2 µL of DNA, and 0.5 µL of Taq DNA polymerase, with deionized water added to a final volume of 25 µL. Amplification of the GH gene and Sanger sequencing were performed according to the manufacturers' protocols (Evrogen and GenSeq, respectively) (Table 1).

Capillary electrophoresis was performed using a Nanofor 05 genetic analyzer with PDMA-6 sequencing polymer and 1× TAPS sequencing buffer (Syntol). Electrophoresis parameters were established according to the GenSeq kit instructions, considering the capillary length (50 cm), polymer type (PDMA-6), and fragment size. Raw sequencing data in AB1 format were analyzed using Mutation Surveyor v5.2.0 (SoftGenetics, Pennsylvania, USA) (Figures 1 and 2).

Blood samples from Kalmyk bull calves (n = 50) were used for genotyping. Blood was collected using vacuum blood collection systems comprising a vacuum syringe-container and specialized needle (Lab-Vac, China). Sampling was performed at the evaluation station in November 2024 (Act No. 5, November 10, 2024), after which the samples were submitted to the Molecular Genetics Laboratory of Kalmyk State University named after B.B. Gorodovikov for analysis.

For physicochemical analysis of beef, longissimus dorsi muscle samples were collected from the right half-carcass between the 9th and 11th ribs following controlled slaughter of bull calves representing the two genotypes (n = 16). Samples were collected from 15-month-old bull calves in June 2025 (Act No. 2, June 7, 2025) and provided to Kalmyk State University named after B.B. Gorodovikov for subsequent transport and laboratory analysis.

Assessment of live weight dynamics, calculation of average daily gain, and physicochemical analyses were conducted without knowledge of genotype-group allocation until completion of the primary analyses. Before the experiment, all operators were trained in standardized weighing procedures, with measurements recorded to the nearest 0.1 kg, and in low-stress animal restraint. Interoperator consistency was maintained at a coefficient of variation <2%. Industrial scales of the NAIS brand (LLC «New Automated Measuring Systems», Russia) were calibrated monthly against reference standards, and calibration results were documented. All experimental procedures were performed using standardized checklists with precise timestamps in accordance with the study protocol.

Materials and reagents included 10% neutral buffered formalin as a fixation solution; hematoxylin, eosin, Van Gieson's solution, and Masson's trichrome for histological staining; reagents required for the respective biochemical analyses; standard buffer solutions; and distilled water. Laboratory equipment included an analytical balance with an accuracy of ±0.0001 g, a spectrophotometer with a wavelength range of 200–1000 nm, a pH meter with an accuracy of ±0.01, a texture analyzer with a load capacity of up to 50 kg, a colorimeter based on the CIE 1976 system, a microtome with a section thickness of 5 µm, and ImageJ software for morphometric analysis.

Analyses were performed according to the corresponding GOST standards: GOST 23042-2015 for determination of fat in meat and meat products; GOST 23041-2015 for determination of hydroxyproline; GOST 25011-2017 for determination of protein; GOST 30178-96 for determination of toxic elements in raw materials and food products by atomic absorption spectrometry; GOST 31727-2012 for determination of total ash; GOST 33424-2015 for determination of magnesium by flame atomic absorption spectrometry; GOST 34132-2017 for determination of the amino acid composition of animal proteins; GOST 55573-2013 for determination of calcium by atomic absorption and titrimetric methods; and GOST R 55484-2013 for determination of sodium, potassium, magnesium, and manganese by flame atomic absorption spectrometry.

Intergroup differences were assessed using Welch's t-test with Bonferroni correction to control the Type I error rate. The adjusted significance level was calculated as α_adj = 0.05/k, where k ranged from 3 to 11 depending on the number of parameters analyzed [32]. Effect sizes were estimated using Cohen's d [33]. For selected analyses, 95% confidence intervals (CIs) for effect sizes were additionally calculated using the noncentral t-distribution [34]. Data distribution was assessed for normality using the Shapiro–Wilk test.

Statistical analyses and data visualization were performed using Python v3.12/3.13 (python.org), pandas v2.2.3 (pydata.org), SciPy v1.14.1 (scipy.org), Matplotlib v3.9.3 (matplotlib.org), and Seaborn v0.13.2 (pydata.org). Results are presented as mean ± standard deviation (M ± SD).

The methodological quality of the study and risk-of-bias were assessed using the Systematic Review Center for Laboratory Animal Experimentation (SYRCLE) risk-of-bias tool in accordance with the Animal Research: Reporting of In Vivo Experiments (ARRIVE) 2.0 guidelines [35, 36]. The assessment covered the principal domains of potential bias, with each domain classified as having a low, moderate, or unclear risk of bias.

The study design and predefined endpoints, including growth performance, slaughter traits, and meat-quality characteristics, were established before initiation of the experiment. Animals were classified according to genotype (LL, n = 37; LV, n = 13) based on objective genotyping results. The genotype groups were comparable in age and initial live weight, and allocation to housing sections was performed using stratified randomization by farm of origin.

Detection bias was minimized through blinding. Personnel responsible for measurements and laboratory analyses were unaware of the animals' genotypes during primary data collection and analysis. Calibrated measurement instruments and standardized experimental procedures were used throughout the study. Attrition bias was considered low because all animals completed the experimental protocol and complete data were available for the predefined outcomes.

Performance bias was minimized by standardizing major husbandry conditions, including diet, housing area (75 m² per section), ventilation, straw bedding, and daily cleaning, together with a 14-day adaptation period. Nevertheless, complete control of microenvironmental conditions was not possible. Potential sources of bias included variations in temperature and humidity, possible fluctuations in feed quality, and microenvironmental gradients related to proximity to feeders and ventilation openings. These factors were not incorporated into the statistical models.

The statistical approach incorporated tests appropriate to the data characteristics, along with Bonferroni correction for multiple comparisons and Cohen's d to estimate effect magnitude. The overall sample comprised 50 animals, with eight animals from each genotype selected for meat-quality assessment. The latter sample size was determined a priori to provide 80% statistical power at α_adj = 0.05 based on an expected Cohen's d > 0.8. The final risk-of-bias assessment is presented in Table 2.

Overall, the study was assessed as having a low risk-of-bias in most domains. The moderate risk of performance bias reflected the inability to completely control microenvironmental conditions within the housing sections. These potential sources of bias were considered when interpreting the findings, while effect-size estimation complemented significance testing by quantifying the magnitude of the observed differences.

The GH c.2141C>G polymorphism in the bull calves (n = 50) was represented by the L and V alleles, with frequencies of 87% and 13%, respectively. The LL and LV genotype frequencies were 74% and 26%, respectively (Figure 3). The VV genotype was not detected in the study population.

The absence of the VV genotype was consistent with the low frequency of the V allele and the correspondingly low expected frequency of VV homozygotes. The V allele was therefore observed exclusively in the heterozygous state in the study population.

Live weight was evaluated in LL and LV bull calves at 8, 12, and 15 months of age (Figure 4). LV calves had numerically higher live weights than LL calves at all three ages. The differences between the LV and LL groups were 4.45 ± 3.5 kg at 8 months, 20.5 ± 6.8 kg at 12 months, and 25.9 ± 6.2 kg at 15 months. However, after correction for multiple comparisons, none of the differences were statistically significant. Cohen's d values ranged from −0.38 to −0.71, indicating moderate-to-large effect magnitudes despite the absence of statistical significance.

Growth intensity was further evaluated by comparing average daily gain between genotypes during the 8–12-, 12–15-, and 8–15-month periods (Figure 5). LV calves had numerically higher average daily gains than LL calves during all evaluated periods. The difference was 11.5% during the 8–12-month period (122 days), 5.1% during the 12–15-month period, and 8.8% over the entire 8–15-month period. None of these differences were statistically significant (p > 0.05). Cohen's d values ranged from −0.20 to −0.58, corresponding to small-to-moderate effect magnitudes. Thus, the present data did not establish a statistically significant association between genotype and average daily gain.

Slaughter traits were evaluated at 15 months of age in eight animals from each genotype-group (Figure 6). Statistically significant differences between the LL and LV groups were detected for carcass weight and internal fat weight (p < 0.05), whereas dressing percentage did not differ significantly between genotypes (p > 0.05). The largest effect-size was observed for internal fat weight (Cohen's d = 3.53; 95% CI = 1.58–5.48), followed by carcass weight (d = 2.28; 95% CI = 1.02–3.54). In contrast, the effect-size for dressing percentage was small (d = 0.14), and its 95% CI included zero.

The chemical composition of beef differed significantly between the LL and LV genotypes for moisture, protein, fat, and ash contents (Figure 7). Beef from LV calves contained, on average, 1.01% more protein and 0.23% more ash than beef from LL calves, whereas LL samples contained 0.50% more fat and 0.40% more moisture. The p-values for these four parameters ranged from <0.001 to 0.0013. Cohen's d values ranged from 1.25 to 1.47, and the corresponding 95% CIs for the mean differences did not include zero, indicating large effect magnitudes.

Essential amino acid composition and amino acid scores were evaluated to characterize the protein quality of beef from the two genotype groups (Figure 8). Mean essential amino acid concentrations were consistently higher in LV than in LL samples, with the largest numerical differences observed for tryptophan (+30%) and histidine (+11%) (Figure 8A).

The amino acid score was calculated as the ratio of the mass fraction of each essential amino acid to that of the reference protein. The analyzed beef samples met the FAO/WHO recommended essential amino acid level of at least 29 g/100 g protein [37]. Amino acid scores exceeded 100% for seven of the nine evaluated amino acids, with valine and isoleucine as the exceptions. High mean amino acid scores were observed for tryptophan and histidine in both genotype groups; however, the differences were not statistically significant after Bonferroni correction (p > 0.0045). Cohen's d values of up to 0.7–0.9 were observed for these amino acids.

No statistically significant genotype-associated differences were detected in the non-essential amino acid composition of the beef samples after correction for multiple comparisons (p > 0.0063) (Figure 9). Moderate effect sizes (Cohen's d up to 0.6–0.8) were observed for glycine and hydroxyproline, with the largest numerical differences occurring for these amino acids. Glutamic and aspartic acids were present at relatively high concentrations in samples from both genotype groups.

The summary analysis of essential, non-essential, and total amino acid contents and their relative proportions is presented in Table 3.

The LV genotype had numerically higher total EAA content (42.28 vs. 40.45 g/100 g protein), total AA content (99.79 vs. 99.02 g/100 g protein), EAA/NEAA ratio (0.74 vs. 0.69), and proportion of EAA in total AA (0.42 vs. 0.41) than the LL genotype. However, none of the parameters presented in Table 3 reached the Bonferroni-adjusted threshold for statistical significance (all p > 0.01). Therefore, these findings indicate numerical genotype-associated differences in the amino acid profile but do not provide statistical evidence of differences for these aggregate amino acid measures under the present study conditions.