1. Introduction
The iron content in cattle muscle tissue is a significant factor in understanding and managing both the health of cattle and the quality of beef produced for consumption. Detecting iron content in the muscle tissue of live cattle typically involves methods that are either noninvasive or minimally invasive, providing a snapshot of the animal’s overall iron status. Blood sample collection and analysis are the most prevalent approaches. This method assesses iron status by measuring parameters such as serum iron, ferritin, and total iron-binding capacity [
1]. While effective, it has its own set of advantages and limitations. Recently, advanced research has shifted toward employing predictive models. These models estimate iron content by considering various factors, including breed, age, diet, and known absorption rates, offering a comprehensive and nuanced understanding of iron content in cattle.
In cattle, alterations in iron homeostasis parameters may occur due to various diseases. For instance, mastitis or parasitic invasions (such as ankylostomiasis) can lead to inflammation, blood loss, and changes in iron metabolism in the animal’s body [
2,
3]. Chronic conditions, such as chronic stress or prolonged periods of poor nutrition, can also affect the iron homeostasis markers in cattle [
4].
Disturbances in iron homeostasis can be reflected in various morphological parameters of blood. Iron deficiency can lead to anemia, which may manifest as a reduction in the hemoglobin (Hb) levels in the blood. Iron deficiency can affect the quantity and shape of red blood cells (RBCs), for instance, resulting in microcytosis (reduced erythrocyte size) and hypochromia (the reduced hemoglobin content in erythrocytes). Additionally, this is accompanied by decreases in Mean Corpuscular Hemoglobin (MCH) and Mean Corpuscular Hemoglobin Concentration (MCHC) [
5,
6,
7,
8]. The determination of transferrin (TF) levels is used in the differential diagnosis of iron-deficiency anemias, characterized by decreased serum iron content, an increase in the level of this glycoprotein, and consequently, a decrease in the percentage of transferrin saturation with iron [
9]. However, in cases of iron overload in the body, the unsaturated iron-binding capacity (UIBC) decreases. Therefore, to assess the blood’s ability to bind iron to transferrin, a laboratory test called total iron-binding capacity (TIBC) is utilized [
10]. The analysis of a soluble transferrin receptor (sTfR) concentration in blood is utilized for investigating iron deficiency anemia and assessing iron the functional status. Due to its insensitivity to inflammatory processes, sTfR can detect anemia in animals already suffering from inflammatory conditions, and it holds particular significance in distinguishing between anemia in chronic conditions and anemia caused by inadequate iron intake [
11,
12]. However, such markers indicate iron deficiency in the overall animal organism and do not provide information about the quantitative level of iron in animal muscle tissue. The method proposed by us will allow for the assessment of iron levels in muscle tissue and, if necessary, its monitoring throughout the animals’ productive use.
The iron content in the muscle tissue of cattle can vary depending on various stochastic and fixed factors, such as the animal’s age, sex, diet, living conditions, breed, and other related aspects [
6,
13,
14]. In this regard, the average iron content in the skeletal muscle tissue of large ruminants varies widely, ranging from 10 to 50 mg/kg [
15,
16,
17,
18,
19,
20,
21,
22,
23,
24,
25,
26,
27,
28]; in earlier studies, iron concentrations of up to 54 mg/kg were found [
29]. A tendency can be noted that, in European studies, the level of zinc in the muscle tissue of animals was typically below 40 mg/kg [
15,
17,
18,
19,
20,
21,
22,
23,
24,
30], except in Slovakia, where samples of muscle tissue from animals raised in the vicinity of the metallurgical plant were studied [
25]. Higher metal concentration levels in skeletal musculature were observed in animals from countries with developing economies [
26,
27,
28].
Approximately 5–15% of the iron ingested through feed is typically absorbed, yet in cases of deficiency, the absorption process can double [
31]. A study conducted by Knowles et al. [
32] demonstrated that the rate of iron reabsorption is inversely proportional to the ferritin level in serum. Iron absorption from the gastrointestinal tract depends on endogenous factors such as age, body iron levels, the gastrointestinal tract environment, and overall health status. Exogenous factors such as chemical form, iron quantity, and other feed components also influence the iron absorption in the intestine, contributing to its increase or decrease. In the in vitro study, it was found that the addition of ascorbic acid increases the absorption of iron from sodium caseinate–ferric iron and ferrous sulfate to a similar level, which significantly surpasses the absorption from iron pyrophosphate [
33]. Phosphorus-containing compounds may be present in the diet of cattle, especially if it includes grain feeds. Phytates and tannins are capable of forming insoluble complexes with iron, thereby hindering its absorption in the cattle’s intestines [
13]. Following erythrocyte breakdown, a significant portion of iron is reabsorbed and utilized for synthesizing new hemoglobin. Erythrophagocytosis occurs in the spleen, liver, and bone marrow. These organs contain siderophages, which phagocytose and degrade old or damaged erythrocytes. The products of erythrocyte degradation, such as hemoglobin, iron, and bilirubin, are then processed and recirculated for further utilization or excretion from the body. Residual iron that has not been absorbed is excreted from the body in feces [
6,
33,
34,
35,
36,
37].
This study is a continuation of exploratory research aimed at developing models for assessing iron levels in muscle tissue.
The models presented in the scientific literature are constantly being improved, providing more accurate and efficient estimates. To understand the focus of this study, we present the key approaches to assessing iron in animal muscle tissue.
Near-infrared spectroscopy (NIRS): NIRS is a nondestructive technique that uses the absorption of near-infrared light by molecules in muscle tissue to estimate the iron content. It is widely used in the livestock industry due to its speed and ease of use. MRI is a noninvasive imaging technique that can provide detailed information about the composition of muscle tissue, including iron content. It offers high-resolution images and is useful for research purposes. Muscle biopsies are commonly used to directly measure iron levels in muscle tissue. These samples can then be subjected to chemical analysis techniques such as atomic absorption spectrometry or inductively coupled plasma–mass spectrometry (ICP–MS) to accurately determine iron concentrations. Ultrasound is used to assess muscle characteristics, including muscle density and fat content. Iron levels cannot be directly measured; however, they can provide valuable information related to the overall quality of muscle tissue. Machine learning models, including artificial neural networks and regression models, can be trained on datasets containing various muscle tissue characteristics, including iron levels. These models can then predict the iron content in muscle tissue based on other measurable parameters, such as age, weight, and breed. Researchers have also explored the use of biochemical markers in blood samples to indirectly assess iron levels in muscle tissue. These markers include serum iron, ferritin, and transferrin saturation levels. Genetic markers associated with iron metabolism and muscle iron content have been identified in cattle. These markers can be used for genetic selection in breeding programs to produce cattle with desired iron levels in their muscle tissue. DXA is primarily used in human medicine but has also been applied to assess the composition of meat samples, including iron content. Moreover, these findings can provide valuable insights into muscle tissue composition.
The choice of model or technique for assessing iron in the muscle tissue of cattle often depends on various factors, such as cost, accessibility, accuracy, and the specific goals of the assessment. Researchers continue to refine these models and explore new technologies to increase the accuracy and efficiency of iron assessments in cattle. These advancements ultimately contributed to improved livestock management and meat quality.
Each method has its own set of advantages and disadvantages. Therefore, providing specialists with more opportunities to assess gland levels in animals can lead to a more effective meat quality. Consequently, our study focused on determining the concentrations of trace elements in biosubstrates. This method assesses the levels of these genes within an organism. Serum blood, hair, and other substances are commonly used as diagnostic indicators [
38,
39]. The mineral contents of biosubstrates differ, which can impact their ability to determine an organism’s elemental status. A drawback of studying mineral substances in serum and blood plasma is that a deficit in these elements appears after the patient becomes symptomatic due to the body’s depletion linked to increased excretion. Therefore, specific changes in the concentrations of individual elements often cannot be detected in a timely manner, and these fluctuations fall within the margins of the error of the analysis method [
40,
41,
42].
Due to the high informativeness of hair in studying the elemental profile, the findings of this research have found broad applications in hygiene, toxicology, and medical investigations, particularly in identifying cases of poisoning by toxic elements [
41,
42,
43,
44,
45].
The primary objective of the present research was to identify an optimal and efficient predictive model for iron levels in the muscle tissue of large ruminants in the Hereford breed. This model aims to assess the animals’ elemental status noninvasively during their lifetime.
2. Materials and Methods
2.1. Ethics Statement
The animals were kept under standard conditions of an industrial complex, complying with veterinary and zootechnical requirements in accordance with legislation [
46,
47] and under standard conditions specific to each species and breed. Feeding was carried out using standard complete compound feed, taking into account the animals’ age, body weight, and productivity. Drinking water for the animals was sourced from local utility and drinking sources, meeting hygiene requirements [
48,
49].
Animal slaughter was conducted in a commercial abattoir in accordance with applicable requirements, technological instructions, and regulatory documents [
50,
51].
2.2. Experimental Design
This study was conducted in 2023 on animals (n = 31) of the Hereford breed raised in the southern region of Western Siberia (Russia).
The sample size was determined for ethical and economic reasons. Conducting research involving the slaughter of farm animals is challenging due to difficulties in accessing and limited resources for data collection. The age of the animals at the time of slaughter was 16–18 months. The animals were raised on a farm located in the south of the Novosibirsk region in the Maslyaninsky district, Russia (coordinates 54°32′45.1″ N 84°13′04.1″ E or 54.545862, 84.217812). The animals were kept on free-range pasture in an ecologically safe area more than 100 km away from industrial enterprises and large cities. The ages of the young bulls is determined by the fact that at 16–18 months, they reached optimal weight, and their physiological growth ends, making it the most economically advantageous time for slaughter.
To search for a model predicting iron levels in the muscle tissue of Hereford cattle, a set of predictors was utilized and subsequently renamed for convenience according to
Table 1. The distribution of the studied characteristics deviated from Gaussian distribution; therefore, to assess the content of each element in the samples under investigation, the median as well as the values of the first and third quartiles (Q1–Q3) were calculated.
Preslaughter health assessments indicate that all the animals were clinically healthy. Rectal thermometry was used to measure the body temperature, which ranged from 37.5 to 39 °C. All animals were fasted for at least 12 h before slaughter and had unrestricted access to water. Samples of skeletal muscle weighing 100 g were taken from the m. obliquus externus abdominis. Muscle tissue and hair samples were collected immediately after slaughter. The selected muscle tissue samples were cooled to 4 °C and dispatched to the laboratory, where they were stored at a temperature of −24 °C until analysis was conducted. Hair samples weighing 10 g for atomic absorption analysis were collected from the withers area. The hair length ranged from 1 to 4 cm. The hair samples were packed in envelopes made of heavy-duty paper and stored frozen, similarly to the muscle tissue.
2.3. Atomic Absorption Analysis
For atomic absorption analysis, a 10 g hair sample was weighed. To clean the hair from impurities, the sample was placed in a flask with distilled water and then mixed for 1 min using a mixer at a rotating speed of 225 RCF. The water was then changed 10 times, repeating this procedure. Subsequently, the hair sample was dyed with acetone (CAS: 67-64-1, XILONG, Shantou, China) and left for 2 min, after which the remaining solution residues were rinsed 3 times with deionized water and dried at room temperature. Then, the hair samples were dissolved in 2 mL of nitric acid (CAS 7697-37-2, XILONG, Shantou, China) and placed in a standard autoclave in the microwave oven MARS-5 (CEM). The autoclave was gradually sealed over 40 min, after which the temperature was increased to 180 °C to perform the dissolution. The resulting solution was transferred to a volumetric flask. The solutions were analyzed after 10- and 100-fold dilutions using calibration solutions prepared based on multielement standards.
The preparation of internal muscle tissue samples for atomic absorption analysis proceeded in the following sequence: the vessel was washed in a soapy solution, rinsed with tap water and then rinsed with bidistilled water before drying. A sample from the test (100 g) was ground using the analytical mill IKA A11 basic and homogenized with an IKA Ultra Turrax Tube Drive control Disperser (IKA-Werke GmbH & Co. KG, Staufen im Breisgau, Germany) until a homogeneous mass was obtained. Subsequently, it was dried in an oven at a temperature of 60–70 °C for approximately 12 h until a constant mass was achieved. From the obtained dry residue, 3 g was weighed and ashed in a muffle electric furnace EKPS 10 (Code 4009) (Smolensk SCTB SPU, Smolensk, Russia) at a temperature of 500–550 °C. After 10–15 h, the mineralization process was completed, and the ash acquired a gray or white color. After cooling to room temperature, the ash residue was dissolved in 3 mL of 50% hydrochloric acid (CAS 7647-01-0, XILONG, Shantou, China) and then dried on a hotplate. This residue was transferred to a volumetric flask and diluted with 25 mL of distilled water [
52]. The concentration of iron in the resulting solution was determined at an analytical wavelength of 510 nm using an analyzer.
The atomic absorption analysis of muscle tissue was conducted using an MGA-1000 spectrometer (Lumex LLC, Saint Petersburg, Russia). The measurements of the chemical element levels in the hair were conducted using the iCAP-6500 spectrometer (Thermo Scientific, Waltham, MA, USA). In the muscle tissue, the concentration of iron was determined, while in the hair, the levels of several heavy metals were determined: P, Ca, Mg, Na, K, Fe, Mn, Cu, Zn, Al, Ba, and Cr.
2.4. Statistical Analysis
Checking the assumptions typical for regression analysis was conducted following the protocol for data exploration to avoid common statistical problems [
53]. Outlier testing for the original data was performed using the Grubbs test [
54]. The assessment of residual distribution normality was executed using the Shapiro–Wilk method [
55]. The detection of related variability between features was carried out using the Spearman correlation coefficient [
56]. The assessment of multicollinearity was performed by calculating the variance inflation factor for each parameter [
57]. The model coefficients were calculated using the method of least squares. Studentized residuals with high Cook’s distance values were analyzed for outliers using Bonferroni correction [
58].
Statistical analysis and visualization of the original datasets were conducted using the R statistical programming language and RStudio development environment.
The use of exploratory analysis in the study was necessary for fitting regression models and selecting a pool of predictors. Initially, an assessment of multicollinearity was conducted. Its presence may render model coefficient estimates unstable, making it challenging to discern the individual contributions of factors to the variance of the response variable. This situation may paradoxically lead to regression model coefficients being statistically insignificant, while the model as a whole remains statistically significant based on the Fisher criterion. Therefore, to assess the associations between variables, Spearman’s correlation coefficients were calculated, and correlation matrices and scatterplots were constructed. Model fitting in the study began with the creation of a full model across all the subsets. Then, as a result of model selection based on internal quality criteria, starting with the full model across all the subsets, two candidate models were selected to check assumptions regarding residuals. To ensure the validity of the models for assessment, the final stage of the study involved checking the assumptions regarding the residuals of the selected model. As multiple regression models are specific examples of general linear models, assumptions regarding residuals align with the Gauss–Markov theorem conditions.