Falah Awwad, Ghassan Al-Sumaidaee, Aya Eltayeb
et al.
ABSTRACT: Fermented dairy products are increasingly valued not only for their nutritional content but also for their potential health-promoting properties. However, assessing these functional benefits often requires time-consuming chemical assays that limit scalability. In this study, we investigated whether deep learning (DL) could offer a faster, more efficient alternative. Using liquid chromatography (LC)-MS quadrupole time-of-flight metabolomics, we analyzed 18 fermented milk samples (derived from camel and bovine milk fermented with different bacterial strains) and measured their bioactivity across 9 in vitro assays, including antioxidant capacity (2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid), 2,2-diphenyl-1-picrylhydrazyl), enzyme inhibition (angiotensin-converting enzyme, degree of hydrolysis), and anticancer activity (HT-29, MDAMB). To address the challenge of limited sample size, we implemented a robust preprocessing pipeline including outlier detection, robust scaling, and data augmentation techniques. We trained a one-dimensional convolutional neural network (1D-CNN; DL) architecture with regularization strategies to predict these bioactivity scores from preprocessed LC-MS data. The model achieved strong performance with a mean absolute error of 0.548 ± 0.089 across all outputs through 3-fold cross-validation, demonstrating effective generalization despite the small dataset. Principal component analysis revealed biologically meaningful structure in the metabolomic data, distinguishing samples by milk type and fermentation condition. Together, these results demonstrate that DL with appropriate regularization and data augmentation can accurately predict the functional bioactivity of fermented milk products from metabolomic signatures, offering a promising path toward scalable, DL-assisted screening in functional food development, even with limited training data.
Due to their markedly distinct protein compositions and structures, goat milk and cow milk display substantially different characteristics. In this study, the quality and composition of goat milk and cow milk were studied after being refrigerated at 4 °C for 7 days, with a particular focus on protein oxidation and aggregation states. The results revealed that alongside increases in acidity, microbial colony count, and hydrolysis, there was a significant change in the protein aggregation state beginning on the second day. This change was characterized by increased turbidity, an elevated centrifugal sedimentation rate, and a right-shifted particle size distribution. After seven days of refrigeration, the centrifugal sedimentation rate of goat milk increased from 0.53% to 0.97%, whereas that of cow milk rose from 0.41% to 0.58%. The degree of aggregation was significantly greater in goat milk compared to cow milk. Additionally, both protein and lipids exhibited substantial oxidation, with the degree of oxidation more pronounced in goat milk than in cow milk. The malondialdehyde (MDA) content increased from 0.047 μg/mL to 0.241 μg/mL in goat milk and from 0.058 μg/mL to 0.178 μg/mL in cow milk. The results suggest that goat milk was more prone to oxidation, which further reduced its stability. Therefore, in the storage and transportation of dairy products before processing, it is essential not only to monitor sanitary conditions but also to effectively control protein oxidation to enhance the quality of milk processing.
Rudy Sykora, Bishal Barman, Hussein M.H. Mohamed
et al.
ABSTRACT: Bovine milk is a nutritionally rich fluid containing bioactive proteins that support immune function and growth. Traditional thermal pasteurization (72°C for >15 s) ensures microbial safety but degrades heat-sensitive proteins. High-pressure processing (HPP) offers a nonthermal alternative for microbial reduction, yet its effect on protein structure, particularly under applications of multiple pressure cycles, remains underexplored. This study aimed to evaluate the effectiveness of single- and dual-cycle HPP treatments for bacterial inactivation and protein preservation in whole bovine milk and to compare these results with the industry standard—HTST processing. Raw bovine milk samples were inoculated with vegetative pathogens (Listeria monocytogenes, Staphylococcus aureus) or spores (Bacillus cereus, Bacillus subtilis) and treated with varying HPP conditions (350–600 MPa; 4–12 min, at 30°C, for single or dual cycles). Microbial reduction was assessed by standard plate count. Whey protein retention (lactoferrin [LF], IgA, IgG, IgM) was quantified using ELISA and compared with HTST and raw milk controls. Dual-cycle HPP treatments significantly enhanced bacterial reduction compared with single-cycle time equivalents for S. aureus and B. subtilis, but not for L. monocytogenes or B. cereus. Treatments for S. aureus demonstrated 0.6 to 2.5 log reduction increases from single to dual cycles at pressures of 350 to 600 MPa. Although no tested treatments achieved >5-log reductions in sporulated B. subtilis, dual-cycles increased reductions by 1.2 log compared with single-cycle time equivalents. Several conditions achieved >5-log reductions for vegetative pathogens, including 600 MPa, 12 min, single cycle; 550 and 600 MPa, 4 min dual cycle, and 550 and 600 MPa, 6 min, dual-cycle. However, all HPP treatments led to substantial degradation of immunological proteins, particularly LF (53%–84% reduction), IgA (86%–95% reduction), and IgM (81%–98% reduction), with protein retention decreasing as pressure and cycle time increased. High-temperature, short-time processing preserved higher levels of native protein structure across all treatments.
C. Heffernan, T.F. O'Callaghan, R. Fitzgerald
et al.
ABSTRACT: The objective of this experiment was to investigate the effect of concentrate supplement level and type on the milk fat production of grazing dairy cows in early to mid-lactation during a high-risk period for reduced milk fat synthesis. Eighty Holstein Friesian dairy cows averaging (mean ± SD) 55 ± 14 DIM were blocked based on their pre-experimental milk production and parity and randomly assigned to 1 of 5 dietary treatments: a pasture-only (P) control supplemented with 0.27 kg of DM/cow per day of a mineral and vitamin pack (P0); P supplemented with 2 kg of DM/cow per day of an industry-standard concentrate (P2); P supplemented with 4 kg of DM/cow per day of an industry-standard concentrate (P4); P supplemented with 4 kg of DM/cow per day of a concentrate containing 10% sodium hydroxide-treated straw (P4S); and P supplemented with 4 kg of DM/cow per day of a concentrate containing 5% calcium salts of fatty acids (P4F). The experiment consisted of an initial 2-wk covariate period, 1 wk of diet acclimatization, and a 12-wk period of data collection. Concentrate supplement level and type had no effect on milk fat concentration. Increasing the concentrate supplementation level linearly increased milk yield, ECM yield, fat yield, protein yield, lactose yield, and milk solids yield. Cows fed P4F had greater milk yield and lactose yield but lower milk protein concentration compared with cows fed P4 and P4S. Compared with the P4S diet, cows fed the P4F diet had greater milk fat yield and tended to produce greater milk solids yield. Cows fed P4F had lower proportions of de novo and mixed fatty acids (FA), as well as greater proportions of preformed FA compared with cows fed P4 and P4S. Cows fed P2 and P4 increased DM and OM intake compared with cows fed P0; however, cows fed P2 and P4 were similar. The total FA intake of cows fed P4 was greatest (400 g/d), cows fed P2 was intermediate (370 g/d), and cows fed P0 was lowest (330 g/d). Changing the concentrate type had no effect on the intakes of total DM, pasture DM, and OM. These results suggest that, although concentrate level and type can affect milk fat yield, they do not affect the milk fat concentration of grazing dairy cows within the conditions investigated in this experiment. Further research is required to determine the nutritional and non-nutritional factors responsible for reducing milk fat concentration in pasture-based systems during the high-risk period.
ABSTRACT: The size of fat globules in ruminant milk to some extent affects the nutritional quality of dairy products and plays potential roles in infant and adult health. Lipid droplets (LD) in mammary epithelial cells are the precursors of milk fat globules (MFG). However, it is unclear what happens to proteins during the transformation process from LD to MFG, and little is known about the regulation of LD diameter in vivo. In this study, 12 mid-lactation Saanen dairy goats were randomly divided into 2 groups: a control group fed a basal diet and an experimental group fed a basal diet supplemented with CLA at 90 g/d. Goat milk was collected for analysis of composition and MFG size. Mammary gland tissue was collected for analysis of LD diameter and proteins. The size of MFG was found to depend on LD diameter in the mammary glands of dairy goats. The regression equations for MFG size (Y) and LD diameter (x) were YD[3,2] = 1.8776x − 1.1984 (R2 = 0.7765) and YD[4,3] = 2.4898x − 0.4453 (R2 = 0.7693), respectively. Proteomic analysis revealed increased expression of proteins associated with lipid droplet autophagy and lipolysis (such as ATG5, ATG7, LDHA, MGL), whereas expression decreased for proteins involved in lipid synthesis (such as PLIN1, FASN, LPL, SCD, APOH). These LD proteins regulated LD diameter through glucose and lipid pathways metabolism, thereby affecting MFG size. Overall, these findings provide the first evidence that LD diameter determines MFG size and highlight the regulatory functions of LD protein in milk fat production.
Thilo von Neumann, Christoph Boeddeker, Marc Delcroix
et al.
The predominant metric for evaluating speech recognizers, the Word Error Rate (WER) has been extended in different ways to handle transcripts produced by long-form multi-talker speech recognizers. These systems process long transcripts containing multiple speakers and complex speaking patterns so that the classical WER cannot be applied. There are speaker-attributed approaches that count speaker confusion errors, such as the concatenated minimum-permutation WER cpWER and the time-constrained cpWER (tcpWER), and speaker-agnostic approaches, which aim to ignore speaker confusion errors, such as the Optimal Reference Combination WER (ORC-WER) and the MIMO-WER. These WERs evaluate different aspects and error types (e.g., temporal misalignment). A detailed comparison has not been made. We therefore present a unified description of the existing WERs and highlight when to use which metric. To further analyze how many errors are caused by speaker confusion, we propose the Diarization-invariant cpWER (DI-cpWER). It ignores speaker attribution errors and its difference to cpWER reflects the impact of speaker confusions on the WER. Since error types cannot reliably be classified automatically, we discuss ways to visualize sequence alignments between the reference and hypothesis transcripts to facilitate the spotting of errors by a human judge. Since some WER definitions have high computational complexity, we introduce a greedy algorithm to approximate the ORC-WER and DI-cpWER with high precision ($<0.1\%$ deviation in our experiments) and polynomial complexity instead of exponential. To improve the plausibility of the metrics, we also incorporate the time constraint from the tcpWER into ORC-WER and MIMO-WER, also significantly reducing the computational complexity.
Linear Recurrent Neural Networks (linear RNNs) have emerged as competitive alternatives to Transformers for sequence modeling, offering efficient training and linear-time inference. However, existing architectures face a fundamental trade-off between expressivity and efficiency, dictated by the structure of their state-transition matrices. Diagonal matrices, used in models such as Mamba, GLA, or mLSTM, yield fast runtime but have limited expressivity. To address this, recent architectures such as DeltaNet and RWKV-7 adopted a diagonal plus rank--1 structure, which allows simultaneous token and channel mixing, improving associative recall and, as recently shown, state-tracking when allowing state-transition matrices to have negative eigenvalues. Building on the interpretation of DeltaNet's recurrence as performing one step of online gradient descent per token on an associative recall loss, we introduce DeltaProduct, which instead takes multiple ($n_h$) steps per token. This naturally leads to diagonal plus rank--$n_h$ state-transition matrices, formed as products of $n_h$ generalized Householder transformations, providing a tunable mechanism to balance expressivity and efficiency. We provide a detailed theoretical characterization of the state-tracking capability of DeltaProduct in finite precision, showing how it improves by increasing $n_h$. Our extensive experiments demonstrate that DeltaProduct outperforms DeltaNet in both state-tracking and language modeling, while also showing significantly improved length extrapolation capabilities.
Jean C. S. Lourenço, Isabela F. Carrari, Georgia C. de Aguiar
et al.
The objectives of this study were to evaluate the effects of supplementing the diet of high-producing Holstein cows with 2-hydroxy-4-(methylthio)-butanoate (HMTBa) on their milk production and composition, milk fatty acid profile, blood metabolites, and body parameters. The study was conducted in a commercial dairy herd in Paraná State, Southern Brazil. One hundred and fifty-eight multiparous cows were used in a randomized block design during 42 experimental days. Cows were distributed into two treatments: the control treatment cows received 100 g/cow/day of corn meal, while the HMTBa-supplemented cows received 35 g of HMTBa + 65 g/cow/day of corn meal. HMTBa supplementation did not alter milk production but improved milk fat content. Cows receiving HMTBa supplementation showed an increase in the concentration of milk medium-chain fatty acids. Serum levels of blood urea and aspartate aminotransferase were lower in HMTBa-supplemented cows. Cows supplemented with HMTBa increased their body condition score. In summary, HMTB supplementation in high-producing Holstein cows improved productive performance, particularly increased milk fat content, altered milk fatty acid profile, and changed some blood metabolites. Our findings contribute to our understanding of using a methionine analogue as a dietary strategy for optimizing milk quality in high-producing Holstein cows.
Arshy Prodyanatasari, Mely Purnadianti, Krisnita Dwi Jayanti
Poor nutrition can cause stunting in children, where children grow short. Blimbing Village is one of the villages in Mojo District, Kediri Regency that raises dairy cows, but the cow's milk products have not been processed optimally. Based on the above, an effort is made to make nutritional value products made from fresh milk and can increase the selling value of dairy products. This PkM activity was carried out through five stages, namely giving pretest, education, demonstration, product tester, and post-test. Education was given about the benefits, processing, and storage of fresh milk and a demonstration of making Silky Strawberry Milk. At the end of the PkM activity, the Silky Strawberry Milk product that had been made was distributed to the participants. Silky Strawberry Milk can be an alternative dessert to meet children's nutritional needs and prevent stunting and become an alternative business to improve the economy of the community and village MSMEs. This PkM activity can increase the knowledge of participants, this is known from the increase in the average post-test score obtained, which is 94.22. Nutrasetics product innovations that utilise natural resources based on local wisdom need to be improved. This is done as an effort to manage, empower, and increase the economic value of these natural resources. In addition, nutrasetika innovation products can fulfil the fulfilment of nutrition and nutrition of children during growth, thus minimising the occurrence of stunting.
ABSTRACT: This article summarizes the applications of biosensors and biomimetic sensors in the detection of residues in dairy products. Biosensors use biological molecules, such as enzymes or antibodies, to detect residual substances in dairy products, demonstrating high specificity and sensitivity. Biomimetic sensors, inspired by biosensors, use synthetic materials to mimic biological sensing mechanisms, enhancing stability and reproducibility. Both sensor types have achieved notable success in detecting pesticide residues, veterinary drugs, bacteria, and other contaminants in dairy products. The applications of biological and biomimetic sensors not only improve the efficiency of residue detection in dairy products but also have the potential to reduce the time and cost of traditional methods. Their specificity and high sensitivity make them powerful tools in the dairy industry, thus contributing to ensuring the quality and safety of dairy products and meeting the growing consumer demands for health and food safety.
Carmen Daniela Petcu, Dana Tăpăloagă, Oana Diana Mihai
et al.
Consumers are increasingly showing in maintaining a healthy dietary regimen, while food manufacturers are striving to develop products that possess an extended shelf-life to meet the demands of the market. Numerous studies have been conducted to identify natural sources that contribute to the preservation of perishable food derived from animals and plants, thereby prolonging its shelf life. Hence, the present study focuses on the identification of both natural sources of antioxidants and their applications in the development of novel food products, as well as their potential for enhancing product shelf-life. The origins of antioxidants in nature encompass a diverse range of products, including propolis, beebread, and extracts derived through various physical–chemical processes. Currently, there is a growing body of research being conducted to evaluate the effectiveness of natural antioxidants in the processing and preservation of various food products, including meat and meat products, milk and dairy products, bakery products, and bee products. The prioritization of discovering novel sources of natural antioxidants is a crucial concern for the meat, milk, and other food industries. Additionally, the development of effective methods for applying these natural antioxidants is a significant objective in the food industry.
Tugce Aydogdu, James A. O’Mahony, Noel A. McCarthy
The ability to measure and capture real-time unit operational data has significant benefits during dairy processing, whether it is the basics, such as measuring temperature, pressure, and flow rates, or more recent developments in the case of in-line viscosity and product-compositional measurements. This rapid data collection has helped increase profitability by reducing energy costs, minimizing product loss, and allowing automated control. Advances in technology have allowed for in-line measurements of the composition and some physical attributes such as particle size and viscosity; however, an attribute that spans both compositional and physical attributes is pH, directly influenced by composition but also environments, such as temperature and dry matter content. pH is measured for a plethora of reasons, such as a measure of milk quality (microbial spoilage), acidification of casein, cheese production, maintaining optimum conditions during protein hydrolysis, etc. However, very little is published on the fundamentals of pH and pH measurement in dairy processing; rather, it is usually a cause-and-effect phenomenon. This review visits one of the oldest analytical considerations in the dairy industry and re-examines how it is affected by product composition and processing conditions.
In this paper, we critically examine the prevalent practice of using additive mixtures of Matérn kernels in single-output Gaussian process (GP) models and explore the properties of multiplicative mixtures of Matérn kernels for multi-output GP models. For the single-output case, we derive a series of theoretical results showing that the smoothness of a mixture of Matérn kernels is determined by the least smooth component and that a GP with such a kernel is effectively equivalent to the least smooth kernel component. Furthermore, we demonstrate that none of the mixing weights or parameters within individual kernel components are identifiable. We then turn our attention to multi-output GP models and analyze the identifiability of the covariance matrix $A$ in the multiplicative kernel $K(x,y) = AK_0(x,y)$, where $K_0$ is a standard single output kernel such as Matérn. We show that $A$ is identifiable up to a multiplicative constant, suggesting that multiplicative mixtures are well suited for multi-output tasks. Our findings are supported by extensive simulations and real applications for both single- and multi-output settings. This work provides insight into kernel selection and interpretation for GP models, emphasizing the importance of choosing appropriate kernel structures for different tasks.
Aymeric Dieuleveut, Gersende Fort, Eric Moulines
et al.
Stochastic Approximation (SA) is a classical algorithm that has had since the early days a huge impact on signal processing, and nowadays on machine learning, due to the necessity to deal with a large amount of data observed with uncertainties. An exemplar special case of SA pertains to the popular stochastic (sub)gradient algorithm which is the working horse behind many important applications. A lesser-known fact is that the SA scheme also extends to non-stochastic-gradient algorithms such as compressed stochastic gradient, stochastic expectation-maximization, and a number of reinforcement learning algorithms. The aim of this article is to overview and introduce the non-stochastic-gradient perspectives of SA to the signal processing and machine learning audiences through presenting a design guideline of SA algorithms backed by theories. Our central theme is to propose a general framework that unifies existing theories of SA, including its non-asymptotic and asymptotic convergence results, and demonstrate their applications on popular non-stochastic-gradient algorithms. We build our analysis framework based on classes of Lyapunov functions that satisfy a variety of mild conditions. We draw connections between non-stochastic-gradient algorithms and scenarios when the Lyapunov function is smooth, convex, or strongly convex. Using the said framework, we illustrate the convergence properties of the non-stochastic-gradient algorithms using concrete examples. Extensions to the emerging variance reduction techniques for improved sample complexity will also be discussed.
ABSTRACT: Free exopolysaccharide (f-EPS) produced by Streptococcus thermophilus improves the texture and functionality of fermented dairy foods. Our previous study showed a major improvement in f-EPS production of Strep. thermophilus 937 by increasing the concentrations of histidine, isoleucine, and glutamate to 15 mM in an optimized chemically defined medium. The aim of this study was to elucidate the effect of His, Ile, and Glu on the growth, f-EPS biosynthesis pathway, and carbohydrate metabolism profiles of Strep. thermophilus 937. The growth kinetics; transcript levels of key genes in the EPS biosynthesis pathway; enzyme activity involved in sugar nucleotide synthesis; concentrations of lactic acid, lactose, and galactose; and extracellular and intracellular pH were analyzed in chemically defined media with different initial histidine, isoleucine, and glutamate concentrations. The results showed that f-EPS production and viable cell counts of Strep. thermophilus 937 increased 2-fold after the concentrations of His, Ile, and Glu were increased. Additionally, increasing the concentrations of His, Ile, and Glu upregulated transcription of EPS biosynthesis genes and increased the activity of key enzymes in sugar nucleotide synthesis. Moreover, the consumption of lactose increased and secretion of galactose decreased, indicating that increasing the concentration of His, Ile, and Glu could enhance f-EPS production by maintaining viable cell counts, promoting sugar nucleotide synthesis, and increasing the transcript levels of the eps gene cluster. Our results provide a better understanding of the effect of AA on EPS biosynthesis in Strep. thermophilus.
ABSTRACT: Although puerperal metritis (PM) is a common infectious disease in dairy cattle, there are currently discrepancies between clinical case definitions within and between available peer-reviewed literature and on-farms practices. The inconsistent use of PM criteria across studies and on-farms practices can result in disparities related to recommendations for treating cows, affecting judicious use of antimicrobials. The objective of this study was to systematically review the peer-reviewed literature for clinical signs used for case definition of PM. The criteria used included local (e.g., vaginal discharge) and systemic clinical signs of infection (e.g., fever, drop in milk). The Preferred Reporting Items for Systematic Review and Meta-Analysis extension for scoping reviews protocols were used to screen commonly used databases. Following this protocol, one reviewer screened title and abstract for eligibility (n = 2,096), followed by full-text screening of selected articles (n = 396) by 2 reviewers to confirm eligible articles (n = 174). The most frequently cited reference article (37.5%) for the definition of PM was published in 2006, followed by articles published between 1998 and 2009 (13%). In 40.2% of articles, no reference was provided for definition of PM; vaginal discharge was described in terms of color, odor, and viscosity when related to the PM definition. Terms used for description of vaginal discharge color were red-brown (61.4%), red (5.1%), brown (8.6%), chocolate (4%), white (1.7%), yellow (0.5%), pink (5.7%), or gray (0.5%); vaginal discharge color was not reported in 24.1% articles. The vaginal discharge odor was described as fetid (75.8%), putrid (5.1%), foul (10.3%), or other (5.7%; e.g., abnormal, malodorous, odoriferous); odor was not mentioned in 7.4% of articles. The vaginal discharge viscosity was described as watery (74.1%), purulent (27%), mucopurulent (8.6%), thin (4%), serous (2.8%), or abnormal (2.3%) and was not mentioned in 11.5% of articles. Fever was included in 59.7% of articles as a criterion for PM diagnosis. The most used rectal temperature threshold was ≥39.5°C (56.8%), followed by ≥39.2°C (2.8%). Approaches used for vaginal discharge evaluation included rectal palpation (37.3%), intravaginal exploration with a gloved hand (18.4%), Metricheck (9.8%), or speculum (5.7%); and in 28.7% of articles, diagnostic tools used were not mentioned. Many of the color and odor vaginal discharge descriptions observed in the literature, used synonymous words to describe the same vaginal discharge sample, highlighting a lack of terminology consensus that could result in disagreements, especially due to the subjective character of these clinical evaluations of vaginal discharge color and odor. Although select consensus articles are available, it is common for studies to disregard a reference when defining PM cases. Furthermore, our findings highlight the need for a robust and clear consensus on criteria and terminology used to diagnose PM.
ABSTRACT: The objectives were to investigate whether supplementation with rumen-protected choline (RPC) during late pregnancy in Holstein cows affects offspring immunity and growth, and whether effects are utero-placental, colostrum dependent, or both. A total of 105 multiparous Holstein cows were assigned randomly to a prepartum diet (1.54 Mcal of NEL/kg of DM, and 15.8% CP) without (control) or with added RPC (12.9 g/d of choline ion). Calves (n = 111) were blocked by sex and assigned randomly to colostrum from control cows or colostrum from RPC cows, resulting in 4 treatments in a 2 × 2 factorial arrangement: (1) calves born and fed colostrum from non-supplemented dams (NN; n = 33); (2) calves from non-supplemented dams and fed colostrum from RPC-fed cows (NC; n = 25); (3) calves from RPC-supplemented dams and colostrum from non-supplemented cows (CN; n = 28); and (4) calves from RPC-supplemented dams and colostrum from RPC-fed cows (CC; n = 25). Growth, intakes, and immunity of females were evaluated up to 56 d of age. Growth and intake of male calves was evaluated up to 35 d of age, and physiological and immune responses to intravenous LPS challenge were evaluated from 21 to 35 d of age. Effects of prenatal and colostrum treatments and interactions between treatments were analyzed using mixed models. Calves fed colostrum from RPC-supplemented dams had a 17.4% increase in apparent efficiency of absorption of IgG compared with calves fed colostrum from control dams (27.4 vs. 23.3%). Incidence of fever in the first 21 d of age tended to be less in females born from RPC-supplemented dams compared with females born from control dams (31 vs. 58%). Prenatal RPC females had increased hematocrit and concentrations of red blood cells, leukocytes, neutrophils, and lymphocytes in blood compared with prenatal females born from control dams. Compared with prenatal control females, prenatal RPC females had greater intake of milk replacer (704 vs. 748 ± 9.9 g/d) and starter (45.4 vs. 60.2 ± 5.9 g/d) during the first 21 d of age. In male calves, mean intake of DM was greater (1,074 vs. 976 ± 45 g/d) after the LPS challenge (0 to 8 d) by calves born from dams fed RPC compared with males born from control dams. Calves born from RPC-fed dams also had lower mean rectal temperature (39.0 vs. 39.2°C) and mean respiration rate (35.6 vs. 39.3 breaths/min) compared with males born from control dams. Moreover, serum concentrations of metabolites (i.e., β-hydroxybutyric acid, fatty acids, and glucose), cytokines (i.e., tumor necrosis factor-α) and acute phase proteins (i.e., serum amyloid A) were consistent with less-severe inflammatory response to LPS in males born from dams fed RPC compared with control. Source of colostrum and interaction between prenatal and colostrum treatments had minimal effects on calf responses to LPS. Overall, maternal RPC supplementation during late gestation suggests a positive effect on immunity, in that colostrum from RPC-fed dams increased efficiency of IgG absorption and maternal supplementation with RPC during late gestation, regardless of colostrum source, attenuated responses to LPS.