Effect of β-casein A1 or A2 milk on body composition, milk intake, and growth in Holstein, Simmental, and crossbred dairy calves of both sexes

The aim of this study was to compare the effects of feeding homozygous β-CN A1 or A2 milk on the body composition, milk intake, and growth of German Holstein (GH), German Simmental (GS), and crossbred (CR) dairy calves of both sexes during the first 2 wk of life. A total of 104 calves (n = 54 female, f; and n = 50 male, m) from the breed groups GH (n = 23), GS (n = 61), and crossbred GH × GS (n = 20) were evaluated. Calves were weighed after birth and received colostrum ad libitum. On the second day, calves were alternately housed in pairs in double-igloo systems according to their random birth order and received either A1 milk (n = 52; 27 female and 25 male) or A2 milk (n = 52; 27 female and 25 male). They were offered 7.5 L/d, and the individual actual total milk intake was recorded. Daily energy-corrected milk intake was also calculated based on the milk composition (fat and protein). Fecal scores were recorded daily. On d 15, visceral adipose tissue (VAT) volume was assessed by open magnetic resonance imaging and dual-energy X-ray absorptiometry (DXA). In addition, fat and lean mass (g), as well as bone mineral content (g) and bone mineral density (g/cm 2 ), were determined by DXA. The body composition, milk intake, and growth were similar between the 2 types of milk in the first 2 wk of life. Female calves had more VAT and fat mass, but less lean mass than male calves. GH and CR calves had more VAT and less lean mass than GS calves. Male calves were heavier than female calves after birth and on d 15. The average days with diarrhea and diarrhea occurrence were similar between calves fed A1 and A2 milk and between both sex groups. GS calves presented slightly more days with diarrhea and increased odds of having diarrhea compared with GH calves, not differing from CR.


INTRODUCTION
β-Casein constitutes approximately 33% of the total protein found in bovine milk (McMahon and Brown, 1984).Dairy cattle possess 12 variants of β-CN, with the most common being A1, A2, and B, exhibiting different genotype frequencies (Sebastiani et al., 2020;Hohmann et al., 2021).The A1 and A2 variants are distinguished solely by an amino acid difference at position 67 of the 209-amino acid protein, where A2 has Pro and A1 has His because of a mutation from the A2 variant (Brooke-Taylor et al., 2017;Chitra, 2022).When the A1 variant undergoes enzymatic digestion, it releases a bioactive opioid peptide consisting of 7 amino acids, known as β-casomorphin-7 (β-CM7; Kullenberg de Guadry et al., 2019).This peptide is not digested in the human body; it activates µ-opioid receptors and increases the risk of diseases such as type-1 diabetes, cardiovascular diseases, higher plasma low-density lipoprotein-cholesterol concentrations, as well as neurological and mainly digestive disorders, including abdominal discomfort and diarrhea (Jianqin et al., 2016;He et al., 2017;Kullenberg de Guadry et al., 2019;Woodford, 2021;Chitra, 2022).Ho et al. (2014) reported significantly higher stool consistency scores in individuals consuming A1 β-CN milk compared with A2 milk.Petrat-Melin et al. (2015) observed a higher in vitro digestion rate for A1 milk compared with A2 milk.In their review, Kullenberg de Guadry et al. (2019) concluded that the results from several studies evaluating the influence of A1 milk as a cause of different diseases were inconclusive because of the significant variability in the results.
In addition to the health aspects associated with the intake of A1 and A2 milk in humans, the performance and health of animals must also be considered.Umbach et al. (1985) reported the presence of β-CM7 in the plasma of 20/24 newborn calves after the first milk intake.The authors suggest that β-CM7 may have important effects on gastrointestinal motility.Hohmann et al. (2021) found β-CM7 values 5 times higher in the plasma of dairy calves fed A1 milk.However, calves fed A1 milk exhibited lower fecal consistency scores and lower prevalence of diarrhea compared with those fed A2 milk.The authors also observed higher milk intake and ADG in calves fed A1 milk.Increased milk intake can lead to an increase in body fat content and alterations in body composition (Kristensen et al., 2007;Keogh et al., 2021).The crude protein content in milk replacers and concentrates can also influence the amount of body fat and lean tissue (Blome et al., 2003;Stamey Lanier et al., 2021).Furthermore, variations in body composition can be expected among different breeds under the same conditions.Scholz et al. (2003) reported lower body fat mass and a higher percentage of lean mass in German Holstein (GH) calves compared with German Simmental (GS) and crossbred (CR) GH × GS calves.
To assess body composition, dissection and chemical analysis have traditionally been used; however, these procedures are destructive, expensive, time-consuming, and prone to inaccuracies due to variations in human evaluations (Sanchez et al., 2022).Consequently, they are not possible for in vivo evaluations.Noninvasive imaging methods serve as an alternative for predicting tissue depths, areas, volumes, or distributions of fat, muscle (water and protein), bone mineral content, and bone mineral density in both live animals and carcasses (Scholz et al., 2015).Currently, emerging image-based techniques such as dual-energy X-ray absorptiometry (DXA), magnetic resonance imaging (MRI), computerized tomography, and ultrasound imaging are being applied for body and carcass evaluations (Scholz et al., 2015;Sanchez et al., 2022).Several studies have reported the use of DXA to estimate live body composition in pigs, sheep, and cattle (Scholz et al., 2003;Hampe et al., 2005;Rothammer et al., 2017;Bernau et al., 2020;Weigand et al., 2020).Magnetic resonance imaging studies have also been conducted on live pigs and sheep (Kremer et al., 2013;Bernau et al., 2016Bernau et al., , 2017Bernau et al., , 2018;;Weigand et al., 2020) but not on calves.Therefore, this study represents the first attempt to evaluate a large number of dairy calves using both DXA and MRI techniques to assess their body composition.The aim of this study was to compare the effects of feeding homozygous β-CN A1 or A2 milk on the body composition, milk intake, and growth of GH, GS, and CR dairy calves of both sexes during the first 2 wk of life.

MATERIALS AND METHODS
All procedures performed on animals in this study were approved by the Animal Ethics Committee of the Government of Upper Bavaria under protocol number ROB-55.2-2532.Vet_03-22-20.The experiment was conducted at the Livestock Center Oberschleissheim, Veterinary Faculty of the Ludwig-Maximilians-University of Munich, from September 2022 to June 2023.All calves born in this period were included in the study and no calves were excluded from the study due to illness or death.

Animals, Feeding, and Evaluations
A total of 104 calves (54 females and 50 males), from the breed groups GH, GS, and CR, were used in the experiment.The calves were randomly assigned to one of 2 feeding groups according to their birth order, not balanced by sex and breed.After every 2 calves born, the feed group was changed: (1) an A1 milk diet with homozygous β-CN genotype A1A1 (n = 52; 27 females, 25 males; 7 GH, 31 GS, and 14 CR), or (2) an A2 milk diet with homozygous β-CN genotype A2A2 (n = 52; 27 females, 25 males; 16 GH, 30 GS, and 6 CR).After birth, the calves were weighed to determine their birth body weight (BBW) and received umbilical cord and navel disinfection, as well as a double set of ear tags with an identification number.The calves were individually housed in calf boxes to receive maternal colostrum ad libitum within the first 24 h (at least 10% of the body weight, with IgG >50 mg/mL measured in a colostrometer).The total colostrum intake was measured (offered − leftover).Blood samples were collected from the jugular vein using EDTA-coated tubes on the second day to determine the transfer of passive immunity.Serum was separated and placed on the measuring surface of the digital refractometer to estimate serum total protein (STP; g/L).On the second day, the calves were housed alternately in pairs in double-igloo systems bedded with straw, according to their random birth order, with ad libitum access to water and hay but no access to calf starter.Because the hay intake during their first 2 wk of life is negligibly small, it does not have to be considered in the feeding ratio.The igloos were identified with the respective treatment.The calves received either A1 or A2 milk in an 8-L nipple bucket 3 times daily: 0600 (3 L), 1130 (1.5 L), and 1730 h (3 L).The individual actual milk intake (MI) of the calves was recorded daily.The milk was collected separately from known homozygous A1 or homozygous A2 cows by using the Lely M4USE system.The composition of the milk offered was determined daily by the Lely A3 or A3 next robotic milking system (Lely Deutschland, Waldstetten, Germany).Because we did not balance the nutrient content of the diet between the 2 Kappes et al.: EFFECTS OF β-CASEIN A1 OR A2 MILK ON DAIRY CALVES treatments, we calculated the daily energy-corrected milk intake (DECMI) using the following equation: (0.327 × MI) + (12.95 × F × MI/100) + (7.65 × P × MI/100), where MI = milk intake, F = fat percentage, and P = protein percentage.On d 15, the calves were deprived of the first feeding and weighed to determine the end body weight (EBW) and calculate ADG and BW gain throughout the study period.We determined the feed efficiency, based on BW gain divided by total energy-corrected milk intake (TECMI, sum of DECMI).
Every day during the second feeding time we recorded the fecal score (FS), where 1 = normal, 2 = soft, 3 = runny, and 4 = watery/diarrhea.Scores of 3 and 4 were considered to be indicators of diarrhea (Renaud et al., 2020).Calves with FS 4 and rectal body temperature >39.5°C were offered an electrolyte solution and received an extra treatment of paromomycin sulfate 140 mg/mL oral solution (Parofor 140, Huvepharma NV) for 3 d.The same person consistently measured and recorded all parameters in all calves each day.
Before the MRI and DXA evaluations on d 15, the health status of each calf was assessed.For the imaging evaluations, the calves were lightly sedated with xylazine 2% (0.4 mg/kg i.m.) for induction and ketamine 10% (1-2 mg/kg i.v.) for sedation maintenance.The anesthetic volumes were adjusted based on vital parameters and eyelid reflex to prevent movements during the evaluations.The study ended at 15 d of age because this was the day of dehorning, for which calves must be sedated in Germany.An experienced veterinarian performed all procedures.

MRI Evaluations
To assess the amount of visceral adipose tissue (VAT), an open low-field MRI system (Siemens Magnetom C!) with a magnetic field strength of 0.35 Tesla was used for the scans, previously established for calves as described by Danesh Mesgaran et al. ( 2020) and for pigs as described by Weigand et al. (2020).The imaging examinations were performed by one person according to appropriate previous training.Before each evaluation, an automated quality control process was conducted by the Siemens Magnetom C! system.For the abdominal evaluation, the calves were positioned in a prone position with the forelimbs and hind limbs extended (Figure 1A).The space between the origin of the last rib and the sacral tuberosity was used as an anatomical landmark for consistent positioning, ensuring reproducible and comparable images of the same body region (Figure 1B).Two sequences were employed to cover the entire region of interest.The first sequence, called "ViscFat sequence," encompassed most of the abdomen, and the second sequence, the "Ham sequence," included the remaining part of the abdomen and the pelvic region.The images were evaluated using the synedra View Personal software (version 16.0.0.3,Synedra information technologies GmbH, Innsbruck, Austria) and the FDA-approved Able 3D-Doctor software (Release 4.0; Able Software Corp., Lexington, MA).The synedra software was used to identify the images covering the region of interest.Subsequently, the Able 3D-Doctor software was employed to manually define the fat depots (white mass, Figure 2B) beneath the abdominal wall, around the kidneys, and in the abdominal cavity for every single slice (Figure 2A).The volume of VAT was quantified in cubic centimeters from a 3-dimensional model, based on the slice number, slice thickness, and slice distance (Figure 2C).

DXA Evaluations
Following the MRI scan, the body composition of the calves was measured using DXA.The GE Lunar iDXA scanner was used for the whole-body scan (Figure 3A).
The technical procedures applied for DXA are described in detail in Danesh Mesgaran et al. (2020).In this study, the standard mode of the scanner was employed to obtain information regarding the quantity and percentage of fat and lean mass, bone mineral content (BMC), and bone mineral density (BMD).The duration of the whole-body scan in standard mode was 7.45 min, with a radiation dose of 3 µGy.During the DXA examination, the calves were positioned in a prone position with the forelimbs flexed and hind limbs extended.Before the examination, a quality control check of the iDXA system was performed.The results of body composition and bone mineral measurements were automatically generated after the scan.The android region (origin of the last rib and the sacral tuberosity) was adjusted individually to encompass the same region of interest as in the MRI scans (Figure 3B), ensuring a meaningful comparison of VAT volume between the 2 methods.The obtained results included BW (kg), BMD (g/cm 2 ), BMC (g), fat and lean mass (g) of the whole body, and VAT volume (cm 3 ) and mass (g) in the android region.

Statistical Analyses
The GPower software (Faul et al., 2007;Rasch et al., 2014;Kang, 2021) served as tool for sample-size calculation.A medium Cohen's effect size (Cohen, 1988) based on numerator degrees of freedom of n = 2 combined with a power (1-β) of 0.8 and an α error of 0.05 led to a sample size of n = 100.We assumed that the special effect (milk type) would explain 10% of the error variance (100%) that resulted in Cohen's effect size of f = 0.316.
All traits were automatically tested for normal distribution by the software SAS 9.4 (SAS Institute Inc., Cary, NC) applying the Kolmogorov-Smirnov test.A variance analysis (ANOVA) was performed by using a mixed-model procedure and REML estimation with SAS 9.4 (MIXED Procedure, SAS Institute Inc., Cary, NC).No exclusion of data points was applied.Milk type, sex, milk type × sex, breed, and sex × breed were defined as fixed effects with EBW as covariate and birth date as random effect for all MRI and DXA body composition traits.No covariate was applied for the records of MI, total protein in blood, and abdominal length.In contrast, BBW was used as covariate for feed efficiency and ADG.A modified model was used for the DECMI with milk type, sex, milk type × sex, breed, sex × breed, age, and milk type × age as fixed effects, measurement date as a random effect, and calf as a repeated effect.The significance level in all cases was set to P < 0.05.
Because the daily recorded FS does not follow an ideal normal distribution, we added a logistic regression analysis using again SAS 9.4 (LOGISTIC Procedure) for the daily recorded trait "diarrhea," which was defined as no diarrhea for fecal scores <3 and diarrhea for fecal scores >2.We calculated the odds ratios (OR) for the binary logit model by applying the Fisher's scoring optimization technique containing the fixed effects milk type, sex, and breed, combined with the covariates BBW, DECMI, rectal temperature, and STP.We applied a backward selection of the covariates and kept the model with the lowest Akaike information criterion.In addition, we performed a logistic regression analysis between diarrhea (yes or no) and the DECMI.Finally, we performed a Bland-Altman analysis to compare DXA VAT with MRI VAT by using a combination of SAS procedures as described by Johnson and Waller (2018).

RESULTS AND DISCUSSION
No interaction between feeding groups and sex was observed, only between sex and breed for the variables DXA VAT, fat mass, and BMD (Table 1).The body composition was similar for the 2 feeding groups, and this was possibly related to the similar MI and ADG.Some differences, nonetheless, were expected based on the results of Hohmann et al. (2021), who reported higher daily MI and ADG for calves fed A1 milk.The short treatment period (15 d) may not have been sufficient to observe any differences in body composition between the feeding groups.Consequently, a more extended observation period would be necessary but may be restricted by the limited spatial capacity of the MRI equipment.However, female calves had more MRI VAT (353 ± 16 cm 3 vs.300 ± 18 cm 3 , P < 0.05) and more DXA body fat (3,246 ± 42 g vs. 2,978 ± 49 g, P < 0.001) compared with male calves, respectively.In contrast, male calves had more DXA lean mass than female calves (47,130 ± 59 g vs. 46,825 ± 50 g, P < 0.05).No BMD or BMC differences were observed (Table 1).Scholz et al. (2003) found no differences between male and female calves in fat and lean tissue percentage, BMD, and BMC between d 6 and d 50 of age; only the BW differed significantly with an advantage for male calves (54.68 ± 0.72 kg vs. 51.52 ± 0.66 kg).A et al. ( 2013) claimed that heifers and castrated males have fatter carcasses, and young bulls produce carcasses with less external and internal fat and more lean mass.These effects are related to the hormonal status of the animals, even in very young animals.Keogh et al. (2021) reported higher systemic concentrations of insulin-like growth factor 1 and insulin in Angus × Holstein heifers fed a high plane of nutrition compared with heifers fed a moderate plane of nutrition from 3 to 21 wk of life.The authors postulated that these 2 anabolic hormones (insulin-like growth factor 1, insulin) may contribute to VAT development by inducing adipogenesis in pre-adipocytes.
In both MRI and DXA evaluations, CR and GH calves had higher amounts of VAT compared with GS (P < 0.05), although there was no difference in the abdominal size (length: approximately 21 cm).Crossbred and GH calves also had more DXA fat mass, but less DXA lean mass compared with GS calves (P < 0.05; Table 1).Because of differences in morphological traits, carcass condition, and dual purpose breed, fat storage occurs in different locations and amounts in the body.Breeds specialized for milk production tend to store more fat tissue, but breeds specialized for beef production store more energy in form of muscle tissue (Pfuhl et al., 2007).Berry (2021) also claimed that Holstein and dairy × beef crossbreds had more fat cover than beef breeds.Pfuhl et al. (2007) found that Holstein bulls had more visceral fat, higher marbling scores, and more intramuscular fat at 18 mo of age, but Charolais bulls put on 3.4% more carcass meat.The same relationship is found in crosses between specialized dairy breeds and dual-purpose breeds.Scholz et al. (2003), however, reported that the GH calves showed a lower body fat content and higher lean content than CR calves, not differing from GS between 4 and 50 d, and Hampe et al. (2005) reported significantly higher body fat percentages for GS compared with GH between 6 and 50 d of age.The 4 different CR combinations between GS and GH had medium body fat and lean percentages in comparison to the 2 purebred lines GS or GH, respectively.Phillips et al. (2017) conducted a comparative analysis of carcasses from Holstein steers and Holstein × Montbéliarde × Viking Red CR steers at approximately 490 d of age.The study found that the fat percentage in the kidney, pelvic, and heart regions was similar between the 2 breeding lines.Additionally, marbling score and backfat thickness (BFT) were also similar.Blöttner et al. (2011) observed higher BFT in Holstein × Brown Swiss cows compared with Holstein cows throughout the first lactation.In a more recent study, Knob et al. (2021) reported higher BFT in Simmental and R 1 Simmental cows compared with Holstein, R 1 Holstein, and F 1 (firstgeneration offspring of two purebred parental breeds) cows during the first 150 d of lactation.Based on these findings, it can be inferred that different sexes and breeds at different ages exhibit variations in fat-storage patterns.
The gold standard for VAT evaluations is the dissection procedure compared with noninvasive techniques (Abate et al., 1994).However, these destructive procedures do not allow for future evaluations on the calves.In our study, we assumed MRI as the reference method for VAT assessment because it is possible to visualize and define the fat depots in all images.Additionally, MRI was considered as the reference based on the results derived in pigs (Weigand et al., 2020) and on the findings of van der Kooy and Seidell (1993) and Abate et al. (1994) for the measurement of human VAT.Compared with MRI VAT, DXA VAT was overestimated by 46 cm 3 (standard error 17 cm 3 , P < 0.01, as determined by a Bland-Altman method comparison using SAS 9.4 with a Student's ttest).Weigand et al. (2020) also observed overestimated DXA values for VAT in pigs compared with MRI data.The accuracy of whole-body analysis in vivo, however, is severely compromised by the gastrointestinal tract of ruminants, resulting in a reduced relationship between body composition from DXA measurements and reference measures (Scholz et al., 2015).
There was no difference in BBW and EBW between feeding groups and among breeding lines, which was ex- Different letters within rows describe significant differences, with P < 0.05.
1 Groups were fed a milk diet with homozygous β-casein genotype A1A1 or with homozygous β-CN genotype A2A2.CR = crossbred (German Holstein × German Simmental); GH = German Holstein; and GS = German Simmental. 2 The y model included BBW as a covariate.
pected because there was no difference in milk consumption.However, male calves were heavier than the female calves in both BBW (43.5 ± 0.8 kg vs. 39.6 ± 0.9 kg) and EBW (53.3 ± 0.8 kg vs. 49.6 ± 0.7 kg), respectively.Other studies also revealed higher BW for male calves due to distinct sexual dimorphism in cattle (Kertz et al., 1997;Dhakal et al., 2013;Hohmann et al., 2021).Knob et al. (2018) found similar growth rates when comparing Holstein and F 1 Holstein × Simmental calves and heifers.Contrary to our findings, Heins et al. (2010) reported that Holstein × Montbéliarde calves weighed 5 kg more than Holstein calves at birth.Hampe et al. (2005) and Scholz et al. ( 2003) also found higher BW for GS and CR calves with a higher proportion of GS than GH and CR calves with a higher proportion of GH.Hohmann et al. (2021) reported that GH calves were the heaviest, followed by GS and CR calves, both at birth and at 21 d of age.
The DECMI was similar between sexes and breeds or crossbreds, but showed a tendency (P = 0.07) toward a higher value for A1 milk than A2 milk (6.64 ± 0.0 L vs. 6.46 ± 0.0 L, respectively).Hohmann et al. ( 2021) observed a higher daily MI for calves fed A1 milk compared with those fed A2 milk (7.28 ± 0.12 L vs. 6.96 ± 0.11 L, P = 0.02, respectively).The authors hypothesized that A2 milk contains a higher protein content, which could explain the lower MI.However, when we considered the milk energy (DECMI; Figure 4), there was no difference between the feeding groups (P = 0.58).The same observation was made for TECMI, which was also similar among sexes and breeds, with an average of 103.6 L during the experimental period.With similar MI, we observed similar ADG among the feeding groups, sexes, and breeds or crossbreds.Although not significant (P = 0.07), Hohmann et al. (2021) observed a slightly better ADG of about 110 g/d for calves fed A1 milk compared with calves fed A2 milk during 3 wk of evaluation.In our study, A1-fed calves gained 699 ± 27 g/d and A2fed calves gained only slightly less, 670 ± 27 g/d.A1 and A2 calves gained 96 g (±3.2 g) and 94 g (±3.3 g) of liveweight, respectively, per liter of milk.This was expected considering the similar TECMI and weight gain.In contrast to our study, Hohmann et al. (2021) found better feed conversion for A1 milk compared with A2 milk, with 9.2 L/kg gain and 10.5 L/kg gain, respectively.Male calves had better feed efficiency than female calves (100 ± 3.4 g/L vs. 90 ± 3.3 g/L, P = 0.02, respectively).
In our study, mean days with diarrhea were similar between feeding groups (P = 0.22, Table 1).There was no association of feeding treatment with likelihood of diarrhea (OR: 0.785; 95% CI: 0.536-1.15;P = 0.21; Figure 5).This is in contrast with Hohmann et al. (2021), who observed higher mean FS (2.56 ± 0.17 vs. 1.97 ± 0.18) and diarrhea prevalence (10% vs. 6%) in calves fed A2 milk compared with those fed A1 milk.Generally, the occur- rence of diarrhea is negatively associated with DECMI.The DECMI decreased by 36.7% if calves had diarrhea (OR: 0.633; 95% CI: 0.584-0.687;P < 0.0001).A similar relationship was reported by de Paula et al. (2017), who observed lower milk replacer intake during the first 3 wk when calves had more frequent diarrhea.Although all calves had an adequate passive immune transfer (STP >60 g/L) on the second day (Tyler et al., 1996;Wilm et al., 2018), GS calves had slightly more mean days with diarrhea (Table 1) and slightly increased odds of having diarrhea (Figure 5) than GH calves, whereas CR calves did not differ from GH and GS calves (Table 1).Twenty (of 20) CR, 21 (of 23) GH, and 57 (of 61) GS calves had at least 1 d with diarrhea.These differences could possibly be related to the different gastrointestinal microbiota (Paz et al., 2016), which can be influenced by the interaction between the host and microbiota, as well as external factors such as the maternal microbiota, calving, diet, and use of antibiotics (Du et al., 2023).Gastrointestinal microbiota have a substantial effect on animal health and productivity as well as diseases (Klein-Jöbstl et al., 2014;Du et al., 2023).However, some groups of microorganisms, such as Lactobacillus, Bifidobacterium, and Faecalibacterium, have been associated with a decrease in the incidence of diarrhea, increase in BW, and better feed conversion rates (Du et al., 2023).Further studies are needed to compare the gastrointestinal microbiota among different breeds and crossbreds at an early age under different health conditions, especially diarrhea.
Currently, there is a tendency to select animals for A2 β-CN (Juan and Trujillo, 2022), mainly because the adverse effects on human health associated with the A1 β-CN variant (Kullenberg de Guadry et al., 2019;Woodford, 2021;Giribaldi et al., 2022;Gonzales-Malca et al., 2023), as well as the observed positive correlation of the A2 β-CN variant with higher milk and protein yields compared with A1 β-CN (Heck et al., 2009;Gai et al., 2021).However, considering the coagulating capacity, Gai et al. (2021) and Giribaldi et al. (2022) reported a higher incidence of the A2 variant in noncoagulating milk, a longer coagulation time, looser curd formation, and a lower cheese yield.Nevertheless, the A2 variant can be beneficial for specific markets due to its suitability for yogurt production or just as whole milk.Before the targeted selection of a specific milk type, it is crucial to consider various factors encompassing human and ani- mal health and performance, as well as physicochemical, technological, and functional characteristics of the milk.
In future research endeavors, the evaluation period should be extended to encompass the weaning and postweaning stages.This long period will facilitate the observation of potential effects of A1A1 and A2A2 milk on key parameters such as ADG, FS, and body composition assessed by DXA.It is noteworthy that DXA provides ample space for the assessment of calves weighing up to 150 kg, previously established by Danesh Mesgaran et al. (2020).The authors also set a benchmark of 150 kg for the use of MRI in assessing calves.However, our observations have revealed a significant limitation when employing MRI in a prone position for VAT assessment.Specifically, we noted that calves weighing up to 70 kg do not fit within the coil.This constraint compromises the quality of imaging, and in some cases, renders image acquisition entirely unfeasible.This did not compromise the MRI images in our study because the heaviest calf in our study weighed less than 69 kg.

CONCLUSIONS
This is the first study to use MRI analysis to provide insights into the body composition of dairy calves.However, feeding A1 or A2 milk to calves has no effect on body composition, MI, and growth of calves during the first 2 wk of life.Female calves had more visceral fat and total body fat, and male calves had more lean mass.Crossbred and GH calves had a higher percentage of VAT and a lower amount of lean mass compared with GS calves.Milk β-CN type had no effect on health parameters such as days with diarrhea and diarrhea occurrence, but GS calves had slightly more mean days with diarrhea and increased odds of having diarrhea than GH calves.Further studies are needed to investigate the effects of A1 and A2 milk on gastrointestinal microbiota considering different breeds or crossbreds.

NOTES
This study was funded partially by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Programa Institucional de Doutorado Sanduíche no Exterior (PDSE), public notice no.10/2022 (Brazil).All procedures performed on animals in this study were approved by the Animal Ethics Committee of the Government of Upper Bavaria under protocol number ROB-55.2-2532.Vet_03-22-20.The authors have not stated any conflicts of interest.

Figure 1 .
Figure 1.Magnetic resonance imaging examination of the abdomen.(A) Position of the calf for the examination of the abdomen by a Siemens Magnetom C! system.(B) Localizer of the "ViscFat Sequence" with defined area encompassing visceral adipose tissue.Each line represents an axial sectional image of the abdomen starting at the origin of the last rib and the sacral tuberosity.
Figure 2. Evaluation of magnetic resonance (MR) images using the Able 3D-Doctor software.(A) Analysis of all slices in the defined body region.(B) T1-weighted (usually fat = bright signal and other tissue less bright [gray or dark] signal may be acquired by a short time of [sequence] repetition) axial MR image with green boundaries, including the VAT.Fat depots (white mass) beneath the abdominal wall, around the kidneys, and in the abdominal cavity were manually defined for every single slice in panel A. (C) Reconstruction of a 3-dimensional (3D) model from the selected boundaries of VAT.The volume of VAT was quantified in cubic centimeters from the 3D model based on the slice number, slice thickness, and slice distance.
Figure 3. (A) DXA examination.(B) DXA evaluation with enCore software.Mineral tissue (left side), soft tissue (right side).The red lines define the android region (from the origin of the last rib to sacral tuberosity).

Figure 4 .
Figure 4. Least squares means for DECMI (L) of calves receiving A1 or A2 milk during the experimental period of 15 d.Error bars indicate SE of estimation.There is no association of treatment with energy-corrected milk intake (P = 0.1615).A1 and A2 = 2 of the most common genetic variants of β-CN found in dairy cattle.

Figure 5 .
Figure 5. Results of the logistic regression analysis of the daily diarrhea records (0/1; 0 = no diarrhea, 1 = diarrhea) for the classes milk type (homozygous for the A1A1 variant of β-CN versus homozygous for the A2A2 variant of β-CN), sex (f = female, m = male), breed (CBREED; CR = crossbred, GS = German Simmental, GH = German Holstein), combined with the covariates BBW and STP.The model was chosen due to the lowest Akaike information criterion compared with models that included additional covariates such as DECMI and rectal temperature.

Table 1 .
Kappes et al.: EFFECTS OF β-CASEIN A1 OR A2 MILK ON DAIRY CALVES Kappes et al.: EFFECTS OF β-CASEIN A1 OR A2 MILK ON DAIRY CALVES Least squares means (±SE) and significance levels (P-value) for body composition, BW, MI, ADG, feed efficiency, FS, days with diarrhea, and STP 1 Variable