Novel prediction models for hyperketonemia using bovine milk Fourier-transform infrared spectroscopy.

Metabolic diseases driven by negative energy balance in dairy cattle contribute to reduced milk production, increased disease incidence, culling, and death. Cow side tests for negative energy balance markers are available but are labor-intensive. Milk sample analysis using Fourier transform infrared spectroscopy (FTIR) allows for sampling numerous cows simultaneously. FTIR prediction models have moderate accuracy for hyperketonemia diagnosis (beta-hydroxybutyrate (BHB) ≥ 1.2 mmol/L). Most research using FTIR has focused on homogenous datasets and conventional prediction models, including partial least squares, linear discriminant analysis, and ElasticNet. Our objective was to evaluate more diverse modeling options, such as deep learning, gradient boosting machine models, and model ensembles for hyperketonemia classification. We compiled a sizable, heterogeneous dataset including milk FTIR and concurrent blood samples. Blood samples were tested for blood BHB, and wavenumber data was obtained from milk FTIR analysis. Using this dataset, we trained conventional prediction models and other options listed above. We demonstrate prediction model performance is similar for convolutional neural networks and ensemble models to simpler algorithm options. Results obtained from this study indicate that deep learning and model ensembles are potential algorithm options for predicting hyperketonemia in dairy cattle. Additionally, our results indicate hyperketonemia prediction models can be developed using heterogeneous datasets.

Dynamics of metabolic characteristics in dairy cows and their impact on disease-free survival time.

Dairy cows are at a greater risk of disease due to increased energy demand during the transition period. Blood biomarkers including beta-hydroxybutyrate (BHBA)1 and non-esterified fatty acids (NEFA)2 are routinely used to identify animals in a state of negative energy balance (NEB)3. Recent research demonstrates cattle have varied response to NEB, that requires multiple blood biomarkers to characterize. This research identified five subcategories (cowtypes) of metabolic responses in transition dairy cows: Healthy, Athlete, Clever, Hyperketonemia, and Poor Metabolic Adaptation Syndrome (PMAS)4. The data set used in this study was collected in Germany by VIT - Vereinigte Informationssysteme Tierhaltung w.V. from 2016 to 2020. Health issues with time of diagnostic were included in the dataset. Using previously reported prediction models for blood BHB and blood NEFA and milk Fourier-transform infrared spectroscopy (FTIR)5 data, the cowtypes in our dataset were predicted. The objective of this study is to evaluate the association of the cowtypes with the disease-free survival time in dairy cows during early post calving using an accelerated failure time regression model. Additionally, transition probabilities of the population dynamics between cowtypes are studied by means of a Markov chain model. Using Healthy cowtype as reference level, Athlete, Clever, and PMAS cowtypes were found to be significant for the disease-free survival probability (P < 0.01). Conversely, Hyperketonemia cowtype was not significant (P = 0.182). Compared to the Healthy cowtype, all other cowtypes had a negative effect on the survival probabilities, which was higher for PMAS cows. Furthermore, after computing the estimated population transition probabilities among cowtypes, the stationary distribution of the Markov chain, along with bootstrap confidence intervals were computed. The results showed 0.091 (95% CI:0.089,0.092), 0.077 (95 % CI:0.074,0.078), 0.684 (95 % CI:0.067,0.069), 0.138 (95 % CI:0.136,0.139), and 0.009 (95% CI:0.008,0.010) of probability of being in Healthy, Athlete, Clever, Hyperketonemia, and PMAS cowtype, respectively. These estimates represent the proportion of cows belonging to the different cowtypes in a herd; information which may prove useful for herd management. The application of blood biomarker predictions using milk FTIR allows us to investigate differences between predicted cowtype and movements between these states and the association with time to disease. Further research will improve our understanding of the dynamic nature of the transition period.

Developing a predictive model for beta-hydroxybutyrate and non-esterified fatty acids using milk fourier-transform infrared spectroscopy in dairy cows.

Negative energy balance following parturition predisposes dairy cattle to numerous metabolic disorders. Current diagnostics are limited by their labor requirements and inability to measure multiple metabolic markers simultaneously. Fourier-transform Infrared spectroscopy (FTIR) data, measured from milk samples, could improve the detection of metabolic disorders using routine milk samples from dairy farms. The objective of this study was to develop a predictive model for numeric values of blood beta-hydroxybutyrate (BHB) and blood non-esterified fatty acids (NEFA). The study utilized a dataset comprised of 622 observations with known blood BHB and blood NEFA values measured concurrently with the milk FTIR data. Using ElasticNet regression on milk FTIR data and production information, we built regression models for numeric blood BHB and blood NEFA prediction and classification models for blood BHB values greater than 1.2 mmol/L and blood NEFA values greater than 0.7 mmol/L. The R2 of the best fitting model was 0.56 and 0.51 for log-transformed BHB and log-transformed NEFA, respectively. The BHB classification model had a 90 % sensitivity and 83 % specificity and the NEFA classification model achieved a sensitivity of 73 % and specificity of 74 %. We applied our numeric prediction models to an external dataset (n = 9660) with known blood metabolites to validate the prediction accuracy of log-transformed blood BHB and log-transformed blood NEFA models. Log-transformed BHB root mean square error (RMSE) was 0.4018 and log-transformed NEFA RMSE was 0.4043. The second objective of this study was to develop a categorization for cows as either metabolically disordered or healthy. Responses to negative energy balance in transition cows are related to blood levels of BHB and NEFA. Cows suffering from metabolic disorders without elevated blood BHB values remain unidentified when detection is focused on blood BHB alone. To account for this differentiated metabolic response, we categorized cows as either 'metabolically healthy' or suffering a 'metabolic disorder' by using a combination of blood BHB, blood NEFA, and milk fat to protein quotient. We obtained a balanced accuracy of 94 % for the prediction of cow metabolic status. Direct prediction of metabolic status can be used to identify hyperketonemic cows in addition to cows exhibiting metabolic response patterns not detected by elevated blood BHB alone.

Full model selection using regression trees for numeric predictions of biomarkers for metabolic challenges in dairy cows.

Dairy cows suffer poor metabolic adaptation syndrome (PMAS)1 during early post-calving periods caused by negative energy balance. Measurement of blood beta-hydroxy butyric acid (BHBA)2 and blood non-esterified fatty acids (NEFA)3 allow early and accurate detection of negative energy balance. Machine learning prediction of blood BHBA and blood NEFA using milk testing samples represents an opportunity to identify at-risk animals, using less labor than direct blood testing methods. Routine milk testing on modern dairies and computer record keeping provide an immense amount of data which can then be used in machine learning models. Previous research for predicting blood metabolites using Fourier-transform infrared spectroscopy (FTIR)4 milk data has focused mainly on individual models rather than a comparison among the models. Full model selection is the process of comparing different combinations of pre-processing methods, variable selection, and statistical learning algorithms to determine which model results in the lowest prediction error for a given dataset. For this project we used a full model selection approach with regression trees (rtFMS)5 . rtFMS uses the cross-validated performance of different model configurations to feed a regression tree for selecting a final model. A total of 384 possible model configurations (algorithms, predictors and data preprocessing options) for each outcome (blood BHBA and blood NEFA) were considered in the rtFMS technique. rtFMS allows direct comparison of multiple modeling approaches reducing bias due to empirical knowledge, modeling habits, or preferences, identifying the model with minimal root mean squared prediction error (RMSE)6 . An elastic net regression model was selected as the best performing model for both biomarkers. The input data for blood BHBA predictions were FTIR milk spectra, with a second derivative pre-processing, and a filter with 212 wave numbers, obtaining RMSE = 0.354 (0.328-0.392). The best performing model for blood NEFA had input data of FTIR milk spectra, with a second derivative pre-processing, and a filter with 212 wave numbers filter along with the time of milking, obtaining RMSE = 0.601 (0.564-0.654). The comparison of multiple modeling strategies, conducted by rtFMS, present an option for improved FTIR prediction models of blood BHBA and blood NEFA by reducing error due to human bias. The implementation of rtFMS to design future prediction models can guide model inputs and features. Our prediction models have the potential to increase early detection of metabolic disorders in dairy cows during the transition period.

A retrospective study investigating the association of parity, breed, calving month and year, and previous parity milk yield and calving interval with twin births in US dairy cows.

Twinning costs the dairy industry an estimated $96 million each year. Twin pregnancy occurrence in high-producing dairy cows is primarily a result of multiple ovulations associated with low circulating concentrations of progesterone due to high milk production. The present retrospective observational study aimed to identify associations between (1) previous parity milk yield and subsequent twin birth prevalence, (2) twin birth with same parity milk production and calving interval (CInt), and (3) twin birth and the subsequent twin calving. The final data set included almost 2.9 million US dairy calving and production records between 2001 and 2020. Variables considered were parity, breed, milk production, CInt, calving month, and year. Logistic and linear regression modeling were used to assess the effects of predictors on outcomes. Herd within state was used as a random effect for all regression models. Twin birth probability increased for cows with increased previous parity milk yield independent of breed or parity. Third and greater parity (3+) compared with second parity (2) and all breeds compared with Jerseys were associated with greater twin probability. Calving between April and September that corresponded to conceiving in July through December was associated with greater twin birth probability. Twin births were associated with decreased milk production following the birth event in Holsteins and parity 2 cows and in the calving months between June and September. Surprisingly, twin births in parity 3+ cows were associated with an increased 305-d milk yield. Cows that had a twin birth were more likely to calve twins in the subsequent parity and had a greater risk of having a CInt between 413 and 600 d. The hazard to subsequent calving after single births was greater compared with twin births. These data can be instrumental in guiding research focus on reducing twinning in lactating dairy cows.

Prediction model optimization using full model selection with regression trees demonstrated with FTIR data from bovine milk.

Predictive modeling is the development of a model that is best able to predict an outcome based on given input variables. Model algorithms are different processes that are used to define functions that transform the data within models. Common algorithms include logistic regression (LR), linear discriminant analysis (LDA), classification and regression trees (CART), naïve Bayes (NB), and k-nearest neighbor (KNN). Data preprocessing option, such as feature extraction and reduction, and model algorithms are commonly selected empirically in epidemiological studies even though these decisions can significantly affect model performance. Accordingly, full model selection (FMS) methods were developed to provide a systematic approach to select predictive modeling methods; however, current limitations of FMS, such as its dependency on user-selected hyperparameters, have prevented their routine incorporation into analyses for model performance optimization. Here we present the use of regression trees as an innovative method to apply FMS. Regression tree FMS (rtFMS) requires the development of a model for every combination of predictive modeling method options under consideration. The iterated, cross-validation performances of these models are then passed through a regression tree for selection of a final model. We demonstrate the benefits of rtFMS using a milk Fourier transform infrared spectroscopy dataset, wherein we build prediction models for two blood metabolic health parameters in dairy cows, nonesterified fatty acids (NEFA) and ß-hydroxybutyrate acid (BHBA). The goal for building NEFA and BHBA prediction models is to provide a milk-based screening tool for metabolic health in dairy cattle that can be incorporated automatically in milk analysis routines. These models could be used in conjunction with physical exams, cow side tests, and other indications to initiate medical intervention. In contrast to previously reported FMS methods, rtFMS is not a black box, is simple to implement and interpret, it does not have hyperparameters, and it illustrates the relative importance of modeling options. Additionally, rtFMS allows for indirect comparisons among models developed using different datasets. Finally, rtFMS eliminates user bias due to personal preference for certain methods and rtFMS removes the dependency on published comparisons of methods. Thus, rtFMS provides clear benefits over the empirical selection of data preprocessing options and model algorithms.

Identifying poor metabolic adaptation during early lactation in dairy cows using cluster analysis.

Currently, cows with poor metabolic adaptation during early lactation, or poor metabolic adaptation syndrome (PMAS), are often identified based on detection of hyperketonemia. Unfortunately, elevated blood ketones do not manifest consistently with indications of PMAS. Expected indicators of PMAS include elevated liver enzymes and bilirubin, decreased rumen fill, reduced rumen contractions, and a decrease in milk production. Cows with PMAS typically are higher producing, older cows that are earlier in lactation and have greater body condition score at the start of lactation. It was our aim to evaluate commonly used measures of metabolic health (input variables) that were available [i.e., blood ß-hydroxybutyrate acid, milk fat:protein ratio, blood nonesterified fatty acids (NEFA)] to characterize PMAS. Bavarian farms (n = 26) with robotic milking systems were enrolled for weekly visits for an average of 6.7 wk. Physical examinations of the cows (5-50 d in milk) were performed by veterinarians during each visit, and blood and milk samples were collected. Resulting data included 790 observations from 312 cows (309 Simmental, 1 Red Holstein, 2 Holstein). Principal component analysis was conducted on the 3 input variables, followed by K-means cluster analysis of the first 2 orthogonal components. The 5 resulting clusters were then ascribed to low, intermediate, or high PMAS classes based on their degree of agreement with expected PMAS indicators and characteristics in comparison with other clusters. Results revealed that PMAS classes were most significantly associated with blood NEFA levels. Next, we evaluated NEFA values that classify observations into appropriate PMAS classes in this data set, which we called separation values. Our resulting NEFA separation values [<0.39 mmol/L (95% confidence limits = 0.360-0.410) to identify low PMAS observations and ≥0.7 mmol/L (95% confidence limits = 0.650-0.775) to identify high PMAS observations] were similar to values determined for Holsteins in conventional milking settings diagnosed with hyperketonemia and clinical symptoms such as anorexia and a reduction in milk yield, as reported in the literature. Future studies evaluating additional clinical and laboratory data, breeds, and milking systems are needed to validate these finding. The aim of future studies would be to build a PMAS prediction model to alert producers of cows needing attention and help evaluate on-farm metabolic health management at the herd level.

Evaluation of the prevalence of digital dermatitis and the effects on performance in beef feedlot cattle under organic trace mineral supplementation.

Bovine digital dermatitis (DD) is a contagious and multifactorial disease that leads to painful, ulcerative lesions of the skin near the heel-horn border of the foot, most commonly in dairy cattle. With regard to beef cattle, the pathogenesis and etiology of DD has not been widely reported or studied over the past several decades. A longitudinal field trial in a commercial feedlot was conducted to compare prevalence and effects of DD in beef steers provided a diet supplemented with a novel formulation of inorganic and organic trace mineral sources (OTM diet) compared to a diet provided with similar levels of trace minerals solely from inorganic sources (CON diet). A secondary objective was to evaluate the prevalence of DD and the potential effects on growth performance and carcass yield and quality. One thousand seventy-seven steers were assigned to 1 of the 2 treatment groups (CON diet or OTM diet) based on location of their home pens which were situated in 1 of 2 barns. All pens in the B barn (group B) were assigned to the OTM diet, and all pens in the A barn (group A) were assigned to the CON diet. The study was conducted in 2 phases: adaptation phase (AP) comprising the initial 60 d of feeding CON and OTM diets and postadaptation phase (PAP) which lasted until cattle were sent to harvest. In the AP, pens in group B had a greater proportion of steers (54.03%) with DD lesions compared to pens in group A (26.72%). During the PAP, the relative risk of observing an increased DD prevalence was significantly ( < 0.05) higher in CON group compared to OTM group. Growth performance, final live weight, and hot carcass weight were negatively impacted when steers were observed to have active DD lesions (M2 lesions) compared to steers with no M2 lesions over the study period. For ADG, a calculated loss per steer of 0.08 kg/d from type I (no M2 lesions) to type II (one M2 lesion; SE = 0.028; = 0.003) and loss of 0.14 kg/d from type I to type III (multiple M2 lesions; SE = 0.038; = 0.0003) were observed. A significant BW loss of approximately 10.06 kg (SE = 4.18; = 0.022) and a mean reduction of 5.5 kg per steer in HCW (SE = 2.74; = 0.043) were also found between type I and type II steers.

Filling gaps in notification data: a model-based approach applied to travel related campylobacteriosis cases in New Zealand.

BACKGROUND: Data containing notified cases of disease are often compromised by incomplete or partial information related to individual cases. In an effort to enhance the value of information from enteric disease notifications in New Zealand, this study explored the use of Bayesian and Multiple Imputation (MI) models to fill risk factor data gaps. As a test case, overseas travel as a risk factor for infection with campylobacteriosis has been examined. METHODS: Two methods, namely Bayesian Specification (BAS) and Multiple Imputation (MI), were compared regarding predictive performance for various levels of artificially induced missingness of overseas travel status in campylobacteriosis notification data. Predictive performance of the models was assessed through the Brier Score, the Area Under the ROC Curve and the Percent Bias of regression coefficients. Finally, the best model was selected and applied to predict missing overseas travel status of campylobacteriosis notifications. RESULTS: While no difference was observed in the predictive performance of the BAS and MI methods at a lower rate of missingness (<10 %), but the BAS approach performed better than MI at a higher rate of missingness (50 %, 65 %, 80 %). The estimated proportion (95 % Credibility Intervals) of travel related cases was greatest in highly urban District Health Boards (DHBs) in Counties Manukau, Auckland and Waitemata, at 0.37 (0.12, 0.57), 0.33 (0.13, 0.55) and 0.28 (0.10, 0.49), whereas the lowest proportion was estimated for more rural West Coast, Northland and Tairawhiti DHBs at 0.02 (0.01, 0.05), 0.03 (0.01, 0.08) and 0.04 (0.01, 0.06), respectively. The national rate of travel related campylobacteriosis cases was estimated at 0.16 (0.02, 0.48). CONCLUSION: The use of BAS offers a flexible approach to data augmentation particularly when the missing rate is very high and when the Missing At Random (MAR) assumption holds. High rates of travel associated cases in urban regions of New Zealand predicted by this approach are plausible given the high rate of travel in these regions, including destinations with higher risk of infection. The added advantage of using a Bayesian approach is that the model's prediction can be improved whenever new information becomes available.

Exploring relationships between Dairy Herd Improvement monitors of performance and the Transition Cow Index in Wisconsin dairy herds.

Transition cow management has been tracked via the Transition Cow Index (TCI; AgSource Cooperative Services, Verona, WI) since 2006. Transition Cow Index was developed to measure the difference between actual and predicted milk yield at first test day to evaluate the relative success of the transition period program. This project aimed to assess TCI in relation to all commonly used Dairy Herd Improvement (DHI) metrics available through AgSource Cooperative Services. Regression analysis was used to isolate variables that were relevant to TCI, and then principal components analysis and network analysis were used to determine the relative strength and relatedness among variables. Finally, cluster analysis was used to segregate herds based on similarity of relevant variables. The DHI data were obtained from 2,131 Wisconsin dairy herds with test-day mean ≥30 cows, which were tested ≥10 times throughout the 2014 calendar year. The original list of 940 DHI variables was reduced through expert-driven selection and regression analysis to 23 variables. The K-means cluster analysis produced 5 distinct clusters. Descriptive statistics were calculated for the 23 variables per cluster grouping. Using principal components analysis, cluster analysis, and network analysis, 4 parameters were isolated as most relevant to TCI; these were energy-corrected milk, 3 measures of intramammary infection (dry cow cure rate, linear somatic cell count score in primiparous cows, and new infection rate), peak ratio, and days in milk at peak milk production. These variables together with cow and newborn calf survival measures form a group of metrics that can be used to assist in the evaluation of overall transition period performance.

Variable selection and regression analysis for the prediction of mortality rates associated with foodborne diseases.

The purpose of this study was to apply a novel statistical method for variable selection and a model-based approach for filling data gaps in mortality rates associated with foodborne diseases using the WHO Vital Registration mortality dataset. Correlation analysis and elastic net regularization methods were applied to drop redundant variables and to select the most meaningful subset of predictors. Whenever predictor data were missing, multiple imputation was used to fill in plausible values. Cluster analysis was applied to identify similar groups of countries based on the values of the predictors. Finally, a Bayesian hierarchical regression model was fit to the final dataset for predicting mortality rates. From 113 potential predictors, 32 were retained after correlation analysis. Out of these 32 predictors, eight with non-zero coefficients were selected using the elastic net regularization method. Based on the values of these variables, four clusters of countries were identified. The uncertainty of predictions was large for countries within clusters lacking mortality rates, and it was low for a cluster that had mortality rate information. Our results demonstrated that, using Bayesian hierarchical regression models, a data-driven clustering of countries and a meaningful subset of predictors can be used to fill data gaps in foodborne disease mortality.

Investigating the genetic background of bovine digital dermatitis using improved definitions of clinical status.

Bovine digital dermatitis (DD) is an increasing claw health problem in all cattle production systems worldwide. The objective of this study was to evaluate the use of an improved scoring of the clinical status for DD via M-scores accounting for the dynamics of the disease; that is, the transitions from one stage to another. The newly defined traits were then subjected to a genetic analysis to determine the genetic background for susceptibility to DD. Data consisted of 6,444 clinical observations from 729 Holstein heifers in a commercial dairy herd, collected applying the M-score system. The M-score system is a classification scheme for stages of DD that allows a macroscopic scoring based on clinical inspections of the bovine foot, thus it describes the stages of lesion development. The M-scores were used to define new DD trait definitions with different complexities. Linear mixed models and logistic models were used to identify fixed environmental effects and to estimate variance components. In total, 68% of all observations showed no DD status, whereas 11% were scored as infectious for and affected by DD, and 21% of all observations exhibited an affected but noninfectious status. For all traits, the probability of occurrence and clinical status were associated with age at observation and period of observation. Risk of becoming infected increased with age, and month of observation significantly affected all traits. Identification of the optimal month concerning DD herd status was consistent for all trait definitions; the last month of the trial was identified. In contrast, months exhibiting the highest least squares means of transformed scores differed depending on trait definition. In this respect, traits that can distinguish between healthy, infectious, and noninfectious stages of DD can account for the infectious potential of the herd and can serve as an alert tool. Estimates of heritabilities of traits studied ranged between 0.19 (±0.11) and 0.52 (±0.17), revealing a tendency for higher values for more complex trait definitions. In terms of genetic selection, all trait definitions identified the best (i.e., most resistant) animals, but only the new trait definitions were able to distinguish between animals with average and high predispositions for DD. Considering repeated measurements resulted in heritability estimates ranging between 0.13 (±0.05) and 0.29 (±0.10).

Survey of facility and management characteristics of large, Upper Midwest dairy herds clustered by Dairy Herd Improvement records.

A survey of management practices was conducted to investigate potential associations with groupings of herds formed by cluster analysis (CA) of Dairy Herd Improvement (DHI) data of 557 Upper Midwest herds of 200 cows or greater. Differences in herd management practices were identified between the groups, despite underlying similarities; for example, freestall housing and milking in a parlor. Group 6 comprised larger herds with a high proportion of primiparous cows and most frequently utilized practices promoting increased production [e.g., 84.4% used recombinant bovine somatotropin (rbST)], decreased lameness (e.g., 96.9% used routine hoof trimming for cows), and improved efficiency in reproduction [e.g., 93.8% synchronized the first breeding in cows (SYNCH)] and labor (e.g., mean ± SD, 67 ± 19 cows per 50-h per week full-time equivalent worker). Group 1 had the best mean DHI performances and followed most closely group 6 for the rate of adoption of intensive management practices while tending to outperform group 6 despite a generally smaller mean herd size (e.g., 42.3 ± 3.6 kg vs. 39.9 ± 3.6 kg of energy-corrected milk production; 608 ± 352 cows vs. 1,716 ± 1,405 cows). Group 2 were smaller herds with relatively high levels of performance that used less intensive management (e.g., 100% milked twice daily) and less technology (33.3 vs. 73.0% of group 1 used rbST). Group 4 were smaller but poorer-performing herds with low turnover and least frequently used intensive management practices (e.g., 39.1% SYNCH; 30.4% allowed mature, high-producing cows access to pasture). Group 5 used modern technologies and practices associated with improved production, yet had the least desirable mean DHI performance of all 6 groups. This group had the lowest proportion of deep loose-bedded stalls (only 52.2% used sand bedding) and the highest proportion (34.8%) of herds not using routine hoof trimming. The survey of group 3 herds did not reveal strong trends in management. The differences identified between herd groupings confirm significant variation in management practices linked to variation in overall herd performance measured by DHI variables. This approach provides an opportunity for consultants and outreach educators to better tailor efforts toward a certain type of dairy management philosophy, rather than taking a blanket approach to applying recommendations to farms simply because of their larger herd size.

First-lactation performance in cows affected by digital dermatitis during the rearing period.

The long-term effects of prepartum digital dermatitis (DD) on first-lactation performance were evaluated in a cohort of 719 pregnant heifers. All heifers were followed for a period of 6 mo until calving and classified on the basis of the number of DD events diagnosed during this period as type I, type II, or type III (no DD, one DD event, and multiple DD events, respectively). Health during the initial 60d in milk (DIM), reproductive and hoof health outcomes, and milk production were compared between the 3 heifer type groups. All logistic and linear models were adjusted for age, height, and girth circumference at enrollment, and the type of trace mineral supplementation during the prepartum period. Overall, cows experiencing DD during the rearing period showed worse production and health outcomes compared with healthy heifers during the first lactation. The percentages of assisted calvings, stillbirths, culled before 60 DIM, and diseased cows during the fresh period were numerically higher in type III cows compared with type I cows. However, none of these differences were statistically significant at the 95% confidence level. Significantly lower conception at first service [odds ratio (OR)=0.55, 95% confidence interval (95% CI): 0.33, 0.89] and increased number of days open (mean=24d, 95% CI: 5.2, 43) were observed in type III cows compared with type I cows. In relation to hoof health, a significantly increased risk of DD during the first lactation was found in type II and III cows (OR=5.16, 95% CI: 3.23, 8.29; and OR=12.5, 95% CI: 7.52, 21.1, respectively), as well as earlier occurrence of DD following calving (OR=59d, 95% CI=20, 96, and OR=74d, 95% CI: 37, 109). Compared with type I cows, statistically significant milk production losses during the initial 305 DIM of 199 and 335kg were estimated in type II and III cows, respectively. This difference was due to a greater rate of production decline (less persistence) after peak yield. No differences in monthly fat and protein percentages or somatic cell counts were observed between the heifer types. Given the long-term effects of DD on health, reproduction, and production, one of the priorities during the rearing period of dairy heifers should be efficient DD prevention and control programs. Such intensive intervention programs based on active long-term DD surveillance, mitigation of risk factors, and prompt treatment are expected to increase overall animal well-being and farm profitability by minimizing the effect of DD during the first lactation.

Shrinking a large dataset to identify variables associated with increased risk of Plasmodium falciparum infection in Western Kenya.

Large datasets are often not amenable to analysis using traditional single-step approaches. Here, our general objective was to apply imputation techniques, principal component analysis (PCA), elastic net and generalized linear models to a large dataset in a systematic approach to extract the most meaningful predictors for a health outcome. We extracted predictors for Plasmodium falciparum infection, from a large covariate dataset while facing limited numbers of observations, using data from the People, Animals, and their Zoonoses (PAZ) project to demonstrate these techniques: data collected from 415 homesteads in western Kenya, contained over 1500 variables that describe the health, environment, and social factors of the humans, livestock, and the homesteads in which they reside. The wide, sparse dataset was simplified to 42 predictors of P. falciparum malaria infection and wealth rankings were produced for all homesteads. The 42 predictors make biological sense and are supported by previous studies. This systematic data-mining approach we used would make many large datasets more manageable and informative for decision-making processes and health policy prioritization.

Cluster analysis of Dairy Herd Improvement data to discover trends in performance characteristics in large Upper Midwest dairy herds.

Principal component analysis (PCA) is a variable reduction method used on over-parameterized data sets with a vast number of variables and a limited number of observations, such as Dairy Herd Improvement (DHI) data, to select subsets of variables that describe the largest amount of variance. Cluster analysis (CA) segregates objects, in this case dairy herds, into groups based upon similarity in multiple characteristics simultaneously. This project aimed to apply PCA to discover the subset of most meaningful DHI variables and to discover groupings of dairy herds with similar performance characteristics. Year 2011 DHI data was obtained for 557 Upper Midwest herds with test-day mean ≥200 cows (assumed mostly freestall housed), that remained on test for the entire year. The PCA reduced an initial list of 22 variables to 16. The average distance method of CA grouped farms based on best goodness of fit determined by the minimum cophenetic distance. Six groupings provided the optimal fitting number of clusters. Descriptive statistics for the 16 variables were computed per group. On observations of means, groups 1, 2, and 6 demonstrated the best performances in most variables, including energy-corrected milk, linear somatic cell score (log of somatic cell count), dry period intramammary infection cure rate, new intramammary infection risk, risk of subclinical intramammary infection at first test, age at first calving, days in milk, and Transition Cow Index. Groups 3, 4, and 5 demonstrated the worst mean performances in most the PCA-selected variables, including DIM, age at first calving, risk of subclinical intramammary infection at first test, and dry period intramammary infection cure rate. Groups 4 and 5 also had the worst mean herd performances in energy-corrected milk, Transition Cow Index, linear somatic cell score, and new intramammary infection risk. Further investigation will be conducted to reveal patterns of management associated with herd categorization. The PCA and CA should be used when describing the multivariate performance of dairy herds and whenever working with over-parameterized data sets, such as DHI databases.

The effect of digital dermatitis on hoof conformation.

Digital dermatitis (DD) is the most prevalent cause of lameness of infectious origin in cattle. However, little is known about the effects of DD on hoof conformation (HC) during the clinical disease. The objectives of the present study were to (1) evaluate the changes in HC observed in feet affected with clinical DD lesions and (2) investigate the temporal relationship between DD and heel horn erosion (HHE). A longitudinal study was carried out including a cohort of 644 Holstein heifers. Digital dermatitis, HC, and presence of HHE in the rear feet of each heifer were assessed during a period of 6 mo. A total of 1,979 feet evaluations were included in the data set, of which 157 corresponded to feet presenting DD lesions >20mm [mean (SD) size of 27.2 (8.2) mm]. Age, days of pregnancy, hip height, and girth circumference were also recorded at cow level. Significant HC changes were observed in DD-affected feet. Results standardized to a period of 90d of follow-up showed an increase in heel height [mean (95% CI) 3.4 (2.5, 4.4) and 2.8 (2.0, 3.7) mm] and claw angle [0.8 (0.2, 1.4) and 1.4 (0.7, 2.0) degrees] of the medial and lateral claws, respectively. In addition, an increase in depth of the interdigital cleft [3.2 (2.7, 3.7) mm] and on debris accumulation [14% (7, 21) of feet] was also observed. Feet affected with clinical DD lesions also experienced a 46% point increase in the presence of severe HHE. In the short term, HC changes returned to normal levels when clinical cure of DD was achieved after topical treatment. In conclusion, significant HC changes occur in heifers affected by clinical DD before lameness symptoms are detected. The transformation of the heel area in feet affected by DD likely promotes the creation of a local environment that favors the persistence of the disease and the occurrence of severe HHE. To avoid further hoof damage, active surveillance and early intervention to reduce HC changes are recommended to improve DD control programs. Successful restoration of HC can be achieved upon clinical cure of DD. The long-term effects in lifetime performance of the HC changes due to DD remain to be further investigated.

Transmission of Escherichia coli O157:H7 in cattle is influenced by the level of environmental contamination.

A pen infection-transmission experiment was conducted to elucidate the role of pathogen strain and environmental contamination in transmission of Escherichia coli O157:H7 (ECO157) in cattle. Five steers were inoculated with a three-strain mixture of ECO157 and joined with five susceptible steers in each of two experimental replicates. Faecal and environmental samples were monitored for ECO157 presence over 30 days. One ECO157 strain did not spread. Transmission rates for the other two strains were estimated using a generalized linear model developed based on a modified 'Susceptible-Infectious-Susceptible' mathematical model. Transmission rates estimated for the two strains (0·11 and 0·14) were similar. However, the rates significantly (P = 0·0006) increased 1·5 times for every 1-unit increase in the level of environmental contamination measured as log10 c.f.u. Depending on the level of environmental contamination, the estimated basic reproduction numbers varied from <1 to 8. The findings indicate the importance of on-farm measures to reduce environmental contamination for ECO157 control in cattle that should be validated under field conditions.

A randomized trial to evaluate the effect of a trace mineral premix on the incidence of active digital dermatitis lesions in cattle.

A balanced, parallel-group, single-blinded randomized efficacy study divided into 2 periods was conducted to evaluate the effect of a premix containing higher than typically recommended levels of organic trace minerals and iodine (HOTMI) in reducing the incidence of active digital dermatitis (DD) lesions acquired naturally and induced by an experimental infection challenge model. For the natural exposure phase of the study, 120 healthy Holstein steers 5 to 7 mo of age without signs of hoof disease were randomized into 2 groups of 60 animals. The control group was fed a standard trace mineral supplement and the treatment group was fed the HOTMI premix, both for a period of 60 d. On d 60, 15 steers free of macroscopic DD lesions were randomly selected from each group for the challenge phase and transported to an experimental facility, where they were acclimated and then challenged within a DD infection model. The same diet group allocation was maintained during the 60 d of the challenge phase. The primary outcome measured was the development of an active DD lesion greater than 20mm in diameter across its largest dimension. No lesions were identified during the natural exposure phase. During the challenge phase, 55% (11/20) and 30% (6/20) of feet were diagnosed with an active DD lesion in the control and treatment groups, respectively. Diagnosis of DD was confirmed by histopathologic demonstration of invasive Treponema spp. within eroded and hyperplastic epidermis and ulcerated papillary dermis. All DD confirmed lesions had dark-field microscopic features compatible with DD and were positive for Treponema spp. by PCR. As a secondary outcome, the average DD lesion size observed in all feet was also evaluated. Overall mean (standard deviation) lesion size was 17.1 (2.36) mm and 11.1 (3.33) mm for the control and treatment groups, respectively, with this difference being driven by acute DD lesions >20mm. A trend existed for the HOTMI premix to reduce the total DD infection rate and the average size of the experimentally induced lesions. Further research is needed to validate the effect of this intervention strategy in the field and to generate prevention and control measures aimed at optimizing claw health based on nutritional programs.

Immune response against Treponema spp. and ELISA detection of digital dermatitis.

The objective of this longitudinal study was to evaluate the immune response against Treponema spp. infection in dairy heifers affected with digital dermatitis (DD). In addition, the accuracy of an indirect ELISA detecting anti-Treponema IgG antibodies in identifying clinical DD status has been assessed. A cohort of 688 pregnant Holstein heifers was evaluated at least 3 times before calving during a period of 6 mo. Complete clinical assessment of DD presence on the back feet of each heifer and blood extraction were performed in a stand-up chute. Digital dermatitis cases were characterized by the M-stage classification system and size and level of skin proliferation. An ELISA was performed on blood serum samples obtained from a subcohort of 130 heifers. For description purposes, the animals were classified by the number of clinical cases experienced during the study period as type I (no clinical cases were observed), type II (only 1 acute clinical case diagnosed), and type III (at least 2 acute clinical cases diagnosed). Multivariable repeated-measures models were used to evaluate the immune response against Treponema spp. infection. A binormal Bayesian model for the ELISA data without cut-point values was used to assess the accuracy of the ELISA as a diagnostic tool. Animals that never experienced a DD event throughout the study kept a constant low level of antibody titer. A 56% increase in mean ELISA titer was observed in heifers upon a first clinical DD case diagnosis. After topical treatment of an acute DD case with oxytetracycline, the antibody titer decreased progressively in type II heifers, achieving mean levels of those observed in healthy cows after a mean of 223 d. Surprisingly, antibody titer was not increased in the presence of M1 (DD lesion <20mm in diameter surrounded by healthy skin) and M4.1 (DD lesion <20mm in diameter embedded in a circumscribed dyskeratotic or proliferative skin alteration) DD stages. Type III cows showed a slight increase in antibody levels. The presence of skin proliferation at first DD diagnosis was found to be associated with an odds ratio of 2.04 of becoming a type III heifer in relation to heifers presenting first lesions without skin proliferation. The ELISA validity was estimated by an area under the curve of 0.88. Predicted probabilities of infection are provided for a range of ELISA values and prevalence of infection. Early detection and treatment is essential to control DD and the ELISA can be used in understanding the immunopathology of DD and shows great promise for prescreening purposes during DD management programs in combination with traditional clinical inspection.