No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Contextual Herd Factors Associated with Cow Culling Risk in Québec Dairy Herds: A Multilevel Analysis
Abstract
Several health disorders, such as milk fever, displaced abomasum, and mastitis, as well as impaired reproductive performance, are known risk factors for the removal of affected cows from a dairy herd. While cow-level risk factors are well documented in the literature, herd-level associations have been less frequently investigated. The objective of this study was to investigate the effect of cow- and herd-level determinants on variations in culling risk in Québec dairy herds: whether herd influences a cow's culling risk. For this, we assessed the influence of herd membership on cow culling risk according to displaced abomasum, milk fever, and retained placenta.
... According to the literature, culling rates in the United States vary from less than 25% to over 35%, with an average of about 30%. In terms of farm profitability, the optimum culling rate should not exceed 30% (Smith et al., 2000); however, in the United States and Canada, culling rates higher than 30% are common (Haine et al., 2017). Studies by Smith et al. (2000) reported culling rates in the United States of 36.3% for what they labeled the "Midsouth region" and 34.5% for the "South region." ...
... Studies by Smith et al. (2000) reported culling rates in the United States of 36.3% for what they labeled the "Midsouth region" and 34.5% for the "South region." More recent studies found annual culling rates in the United States of approximately 32% and 30.3% in Québec, Canada (Haine et al., 2017). In contrast, Pinedo et al. (2010) reported annualized rates of live culling and death to be 25.1% and 6.6%, respectively. ...
... Regardless of the variability in culling rate, excessively low or high removal rates can be seen as a sign of management problems. However, most importantly, the management goals and dynamic of each individual herd must be considered to determine its optimal culling rate (Haine et al., 2017). ...
Although more than 3 million head of dairy cows enter the food supply chain in the U.S. every year, research on this topic remains limited and scarce. Meat production from dairy cows is a significant component of beef production, accounting for almost 10% of U.S. commercial beef production. Thus, the purpose of this review is to demonstrate the importance of dairy cows as a beef source, and to provide an overview on topics from farm to meat product – culling, marketing, transportation, welfare, body composition and its relationship with lactation particularities, carcass characteristics, meat quality, and traceability. Current scientific evidence has shown that culling a dairy cow at an appropriate time has beneficial effects on cow welfare and, consequently, cow value. During marketing the dairy cow is visually evaluated for health and factors associated with its expected carcass value; thus, marketing a well-conditioned cow will ensure that the animal is fit for transportation and provides high carcass yield. However, limitations such as low body condition score, lameness or mobility problems, and visual defects remain persistent. Even though beef harvest plants accommodate cows in all body composition states, the current carcass grade system does not reflect the mature cow industry needs. Therefore, improvement of the grading system could maximize carcass utilization and increase cow carcass value by recognizing subprimal cuts that could be merchandised as whole muscle cuts. Lastly, implementation of a traceability system would unify information from the farm to harvest assisting the industry in making further advancements.
... The identified risk factors mostly belong to the groups of animal factors (e.g. milk yield or occurrence of diseases and success of reproduction), as well as farm housing conditions, feeding and management factors (Chiumia et al., 2013;Haine et al., 2017;Rilanto et al., 2020;Weigel et al., 2003). The genetic trend of productive life is increasing, however simultaneously with improving knowledge and improvements in genetics, cow longevity shows a decreasing trend (De Vries, 2017;Van Pelt et al., 2016). ...
... The distribution of herd CR and longevity measured as the MAofCC were comparable to that reported in many other countries in the last decade (Bergeå et al., 2016;Chiumia et al., 2013;De Vries, 2017;Hadley et al., 2006;Haine et al., 2017;Nor et al., 2014;Schuster et al., 2020). Although analyzing the association between herd parameters (herd size and level of milk yield) and cow culling rates and longevity was not the primary interest of this study these factors were included as background information. ...
The farmer has the central role in determining cow culling policies on their farm and thus affecting cow longevity. The present study aimed to examine farm managers´ satisfaction, attitudes, personality traits and analyse the associations with dairy cow culling and longevity in large commercial dairy farms.
Farm managers of 116 dairy herds rearing at least 100 cows in freestall barns were included. A questionnaire for the farm managers registered personal background information of respondent and included statements capturing their satisfaction, opinions and attitudes regarding dairy cow culling and longevity, farming in general, and a Ten Item Personality Inventory scoring. For each herd, the last 12 months cow culling rate (CR, excluding dairy sale) and herd mean age of culled cows (MAofCC) was obtained from the Estonian Livestock Performance Recording Ltd. A K-mean clustering algorithm was applied to subgroup farm managers based on their attitudes, opinions and personality traits.
The yearly mean herd CR was 33.0% and MAofCC was 60.6 months. Farm managers´ were mostly dissatisfied with cow longevity and culling rates in their farms. Dissatisfaction with culling rates and longevity, priority for producing high milk yields over longevity and production-oriented attitude was associated with high culling rates and poor longevity. Farm managers' personality had an effect on herd culling rates and their attitudes explained one third of the variability of culling rates and longevity.
Explaining the economic consequences of high culling rates and decreased longevity, improving the visibility of these parameters together with benchmarking could bring these issues into focus.
... Nieuwhof et al. (1989) reported an average productive lifespan of 38.4 months for Holsteins that first calved after 1965 in the United States. These productive lifespans are similar to or slightly higher than averages reported for other countries (Mohd Nor et al., 2014;Haine et al., 2017). Within countries, annual cull rates vary greatly, however. ...
... The main driver of these lower cow cull rates points to suboptimal decision-making by dairy farmers, for example, because they may give cows not enough opportunity to get pregnant or underestimate the cost to raise replacement heifers. Furthermore, criteria for culling vary between farmers (Beaudeau et al., 1996) and between farmers and their advisors (Haine et al., 2017). There is a lack of current studies that determine the economically optimal productive lifespan of dairy cows. ...
The average productive lifespan is approximately 3 to 4 years in countries with high-producing dairy cows. This is much shorter than the natural life expectancy of dairy cattle. Dairy farmers continue to cull cows primarily for reasons related to poor health, failure to conceive or conformation problems prior to culling. These reasons may indicate reduced welfare leading up to culling. Improvements in health care, housing and nutrition will reduce forced culling related to these welfare reasons. However, productive lifespan has remained similar in decades, despite large improvements in cow comfort and genetic selection for the ability to avoid culling. On the other hand, genetic progress for economically important traits is accelerating within the last decade, which should slightly shorten the average economically optimal productive lifespan. A major driver of productive lifespan is the availability of replacement heifers that force cows out when they calve. The average productive lifespan could be extended by reducing the supply of dairy heifers, which would also have benefits for environmental sustainability. Improvements in culling decision support tools would strengthen economically optimal replacement decisions. In conclusion, major factors of the relatively short productive lifespan of dairy cows are welfare-related, but other economic factors like supply of heifers, genetic progress and non-optimal decision-making also play important roles.
... Dubuc et al. reported that postpartum diseases frequently occurred in Canadian dairy herds, and alarm levels determined using median herd prevalence of postpartum diseases were identified as risk factors for poor reproductive performance and increased culling [8]. Among postpartum diseases, displaced abomasum and milk fever are known as risk factors for culling [2,8,21,26]. Furthermore, hyperketonemia is known to increase the likelihood of displaced abomasum [11,34]. ...
... Consequently, it negatively affects reproductive performance [10,14,20]. By itself, retained placenta is known to have no effect on culling [9,21,26]; however, some studies point to this condition as a risk factor [2,8]. In the present study, the only independent variable was the occurrence of diseases during the early postpartum period. ...
Peripartum disorders in dairy cows negatively influence their productivity and reproductive performance. However, only a few reports have clearly indicated the influence of such disorders on the productivity and reproductive performance at a local-area or cow-level in Japan. This study aimed to elucidate the influence of diseases occurring within 60 days after calving on subsequent productivity and reproductive performance. Accordingly, a wide-area database on dairy production was used for epidemiological analysis; subsequently, multivariable analysis was performed to investigate the association of such diseases with productivity or reproductive performance in 6,545 cows from 178 farms in Fukuoka. We used 305-day energy-corrected milk (305 ECM) as an index of productivity and conception and culling as indices of reproductive performance. With regard to causality, mixed-effects model was used for analyzing the association between disease and productivity, and Cox proportional hazard model was used for analyzing the association between disease and reproductive performance. Compared to the disease absence group, the disease presence group demonstrated significantly lower 305 ECM [−154 kg; 95% confidence interval (CI), −229 to −79] and risk of pregnancy [hazard ratio (HR), 0.85; 95% CI, 0.80–0.91] and higher risk of culling (HR, 1.36; 95% CI, 1.17–1.59). These results indicate that, in Fukuoka, dairy cows affected by diseases within 60 days after calving exhibit lower productivity and reproductive performance. Therefore, proper dairy cow management during the peripartum period to prevent diseases during early lactation may maintain or improve productivity.
... Culling decisions are made based on factors belonging to the cow, such as health, milk production, and reproductive status, but also on external factors such as the availability of replacement heifers given the herd's reproductive performance, parlor capacity, milk and beef prices, land availability, or even country-specific quotas (De Vries, 2017). As in this study, culling reasons are linked to a specific farming context (Martin , 1982), although a recent study from Canada suggested that the herd effect on culling risk is only minor (Haine et al., 2017c). To our knowledge, this is the first study describing the association between MD and culling in a transition management facility, where preventive strategies are adopted to limit the onset of transition diseases, and it could therefore be useful for future comparison with herds with similar management strategies. ...
... Results regarding RP are not surprising. Many studies reported no significant effect of RP on the culling risk (Dohoo and Martin, 1984;Erb et al., 1985;Dubuc et al., 2011;Haine et al., 2017c). Beaudeau et al. (1995) found that RP was even protective against culling, at least during the first lactation. ...
The objective of this study was to assess the association between individual metabolic diseases (MD) and multiple MD (MD+) in the transition period (±3 wk relative to calving) and the culling risk in the first 120 d in milk (DIM) in Holstein-Friesian dairy cows. Health records from a transition management facility in Germany with 1,946 calvings were analyzed in a 1-yr cohort via survival analysis and a decision tree model. The recorded MD were milk fever (MF), retained placenta (RP), metritis (METR), ketosis (KET), displaced abomasum (DA), twinning (TWIN), and clinical mastitis (MAST). The overall culling within 120 DIM was 18%. The 120 DIM culling risk for healthy cows (64.8% of the total) was 13%, whereas it was 25% for MD (24.5%) and 33% for MD+ (10.7%) cows. The 120 DIM culling risk (%) for each MD and MD+, respectively, was 34.6 and 48 for MF and MF+, 15 and 31.5 for RP and RP+, 9.4 and 22.2 for METR and METR+, 30.7 and 37.3 for KET and KET+, 56.1 and 46.8 for DA and DA+, 30.3 and 34 for TWIN and TWIN+, and 36.6 and 27.8 for MAST and MAST+. Moreover, the incidence risk (%) for each MD and MD+, respectively, was 4 and 2.6 for MF and MF+, 1 and 2.8 for RP and RP+, 8.7 and 6 for METR and METR+, 4.5 and 6.1 for KET and KET+, 0.8 and 2.4 for DA and DA+, 1.7 and 2.7 for TWIN and TWIN+, and 3.6 and 1.8 for MAST and MAST+. Setting the healthy cows as the referent, the 120 DIM hazard ratios (HR) for culling were MD 2.1, MD+ 2.9, MF 3.3, MF+ 4.6, RP+ 2.7, METR+ 1.8, KET 2.6, KET+ 3.3, DA 5.5, DA+ 4.5, TWIN 2.8, TWIN+ 3.0, MAST 3.1, and MAST+ 2.3. According to both decision tree and random forest analyses, MF was the most significant disease influencing survival, followed by DA, MAST, METR, and TWIN. In conclusion, the presence of MD or MD+ during the transition period was associated with increased culling risk in the first 120 DIM. The culling hazard was greater when an MD was complicated with another MD. In this study performed in a well-managed large farm, uncomplicated cases of RP (HR = 1.2) and METR (HR = 0.7) did not have an influence on the 120 DIM culling risk. Interestingly, both decision tree and random forest analyses pointed to MF and DA as main culling reasons in the first 120 DIM in the present dairy herd.
... The risk of culling is not constant over the life of a cow. It depends on cow factors such as lactation number, stage of lactation, milk yield, and reproductive status as well as environmental factors such as season of calving and herd-production needs [17,[19][20][21][22]. Death, as well as diseases and injury, are the main reasons for culling early after the onset of a new lactation [19,20]. ...
The ability of dairy farmers to keep their cows for longer could positively enhance the economic performance of the farms, reduce the environmental footprint of the milk industry, and overall help in justifying a sustainable use of animals for food production. However, there is little published on the current status of cow longevity and we hypothesized that a reason may be a lack of standardization and an over narrow focus of the longevity measure itself. The objectives of this critical literature review were: (1) to review metrics used to measure dairy cow longevity; (2) to describe the status of longevity in high milk-producing countries. Current metrics are limited to either the length of time the animal remains in the herd or if it is alive at a given time. To overcome such a limitation, dairy cow longevity should be defined as an animal having an early age at first calving and a long productive life spent in profitable milk production. Combining age at first calving, length of productive life, and margin over all costs would provide a more comprehensive evaluation of longevity by covering both early life conditions and the length of time the animal remains in the herd once it starts to contribute to the farm revenues, as well as the overall animal health and quality of life. This review confirms that dairy cow longevity has decreased in most high milk-producing countries over time and its relationship with milk yield is not straight forward. Increasing cow longevity by reducing involuntary culling would cut health costs, increase cow lifetime profitability, improve animal welfare, and could contribute towards a more sustainable dairy industry while optimizing dairy farmers’ efficiency in the overall use of resources available.
... LDA has a reported lactational incidence risk of 1.21%-6% [2,3] and is mostly diagnosed within 6 weeks postpartum [4]. The conservative treatments in the forms of rolling of the cow and correction of systemic electrolyte derangements often fail to resolve the condition [5,6], and cattle are most likely to either undergo surgical treatment or to be electively culled with major economic and welfare consequences [6,7]. Recently, LDA was ranked above mastitis, lameness, metritis, retained placenta, ketosis, and hypocalcaemia as a major cause of economic losses in the US dairy industry [8]. ...
... LDA has a reported lactational incidence risk of 1.21%-6% [2,3] and is mostly diagnosed within 6 weeks postpartum [4]. The conservative treatments in the forms of rolling of the cow and correction of systemic electrolyte derangements often fail to resolve the condition [5,6], and cattle are most likely to either undergo surgical treatment or to be electively culled with major economic and welfare consequences [6,7]. Recently, LDA was ranked above mastitis, lameness, metritis, retained placenta, ketosis, and hypocalcaemia as a major cause of economic losses in the US dairy industry [8]. ...
Objectives: Left displaced abomasum (LDA) is a common postparturient condition of high yielding dairy cattle. The diagnosis of LDA is challenging and has historically been based on findings that are not specific to the condition. The objective of the current study was to investigate the diag¬nostic performance of ultrasonography (USG) in the clinical management of dairy cows identified with left-sided ping sound postpartum.
Materials and methods: Cows with reduced appetite postpartum and had audible left-sided ping sounds on abdominal auscultation were eligible to be prospectively recruited onto the study. The results of clinical findings and abdominal USG were recorded along with milk β-hydroxybutyrate levels, pH levels of abomaso/rumenocentesis samples, and findings on exploratory laparotomy. The diagnostic performance of USG and other clinical investigations was assessed by calculating the test sensitivity and specificity using exploratory laparotomy as a gold standard test.
Results: A definitive diagnosis of LDA was made in 23 cows, 8 cows were diagnosed with peritoni¬tis, and 4 cows with frothy tympany. The USG findings that were consistent with LDA were present in all cattle diagnosed with LDA at exploratory laparotomy. The USG findings over the past three intercostal space characteristics of LDA, however, were also present in five cases subsequently diagnosed with peritonitis and in all cases diagnosed with frothy tympany on exploratory lapa¬rotomy. The pH of abdomaso/rumenocentesis samples yielded the highest diagnostic accuracy (97.14%) as a single test in the current study.
Conclusions: USG over the left abdominal wall despite being a highly sensitive test for the diagno¬sis of LDA has limitations as a diagnostic tool due to suboptimal specificity. [J Adv Vet Anim Res 2020; 7(2.000): 308-313]
... Egy több, mint 43 ezer ellést felölelő vizsgálatban a klinikai méhgyulladás 15,2-szer nagyobb valószínűséggel alakult ki a nem MBV-s tehenekhez képest (22). Feltehetően csak azokban az MBV-esetekben csökken a tejhozam és romlanak a szaporodási mutatók, amikor a méhgyulladás is megjelenik (20), a selejtezés esélyét azonban az MBV önmagában nem növeli (15,16). ...
Background: The management of the transition period is of utmost importance for profitable dairy production, because the risk of diseases with potentially large economic effect is high.
Objectives: The aim of this study was to quantify the effect of postpartum uterine diseases on the major reproductive parameters, and to estimate the resulting economic loss.
Materials and Methods: The data of 3,660 calving events that occurred in 2016 and 2017 on five large-scale Hungarian Holstein-Friesian farms were analysed. Information regarding uterine treatments, retained placenta and inflammatory uterine diseases were gathered. The major reproductive parameters (i.e. calving to conception interval – CCI, services per conception – SPC, and first service conception risk – CR1) were calculated based on cow-level data. Statistical analyses were performed by using linear and logistic regression, and Dunnett-test. Losses due to open days, excess semen use and drug cost were taken into account in the economic calculations (1 EUR = 320 HUF).
Results and Discussion: Uterine treatments were performed after 42.68% of the calvings, of which 13.28% were done due to retained placenta, and 29,40% due to uterine inflammation. Uterine treatments and retained placenta were more likely in multiparous cows (odds ratio: 1.22 and 2.05, p = 0.0098 and p < 0.0001, respectively). Retained placenta and uterine inflammations increased CCI by 2.7 and 28.3 days, SPC by 0.9 and 2.2, and reduced CR1 by 4.9 and 4.0 percentage points, respectively. The economic loss due to retained placenta amounted to 38.8 EUR per case, of which treatment cost had the largest share (46.4%). Uterine inflammations caused 122.8 EUR loss per case, with increased number of days open responsible for 57.6% of this loss.
This analysis introduces a stochastic herd simulation model and evaluates the estimated reproductive and economic performance of combinations of reproductive management programs for both heifers and lactating cows. The model simulates the growth, reproductive performance, production, and culling for individual animals and integrates individual animal outcomes to represent herd dynamics daily. The model has an extensible structure, allowing for future modification and expansion, and has been integrated into the Ruminant Farm Systems model, a holistic dairy farm simulation model. The herd simulation model was used to compare outcomes of 10 reproductive management scenarios based on common practices on US farms with combinations of estrous detection (ED) and artificial insemination (AI), synchronized estrous detection (synch-ED) and AI, timed AI (TAI, 5-d CIDR-Synch) programs for heifers; and ED, a combination of ED and TAI (ED-TAI, Presynch-Ovsynch), and TAI (Double-Ovsynch) with or without ED during the reinsemination period for lactating cows. The simulation was run for a 1,000-cow (milking and dry) herd for 7 yr, and we used the outcomes from the final year to evaluate results. The model accounted for incomes from milk, sold calves, and culled heifers and cows, as well as costs from breeding, AI, semen, pregnancy diagnosis, and calf, heifer, and cow feed. We found that the interaction between heifer and lactating dairy cow reproductive management programs influences herd economic performance primarily due to heifer rearing costs and replacement heifer supply. The greatest net return (NR) was achieved when combining heifer TAI and cow TAI without ED during the reinsemination period, whereas the lowest NR was obtained when combining heifer synch-ED with cow ED.
With growing adoption of precision dairy technologies, the use of big data is becoming increasingly common in the dairy industry. The speed at which data are generated has led to increased interest in developing detection and predictive models for animal health and disease events using real time records. When combining data from multiple sources, statistical methods exist to account for the underlying heterogeneity in data collected from commercial farms, although its impact on predictive models is not known. We investigated how 4 different issues commonly seen in these large datasets impact the performance of deep recurrent neural networks (RNNs) trained to detect the onset of clinical mastitis (CM) in dairy cows. Data were simulated by first sampling from real-world data and adding noise, then defining the association between predictor variables and CM while incorporating parameters to reflect underlying heterogeneity: 1) random effects to reflect unmeasured variability at the farm level (3 levels – none, moderate, high); 2) random effects to reflect unmeasured variability at the cow level (3 levels – none, moderate, high); 3) missed recording of CM cases (3 false-negative rates – 0.10, 0.25, 0.50); and 4) incomplete observations due to certain farms not having a somatic cell count sensor (SCC data missing vs SCC data included). At baseline (moderate farm and cow random effects; moderate misclassification; 42% herds with SCC sensor) the model achieved a sensitivity and specificity of 86% and 90% respectively. Higher levels of unmeasured variability at the farm and cow levels resulted in reduced model performance (sensitivity and specificity of 76% and 85% at the highest levels), indicating that data collection and feature selection should be informed by previous knowledge of the associations between the outcome and predictors when possible, and that model performance may be limited when predictors are selected only from routinely collected data. However, even when 50% of CM cases were incorrectly recorded as CM-negative, model performance did not decrease, demonstrating that deep RNNs are robust to the level of misclassification that would be typically encountered in dairy datasets. RNNs were also able to accurately detect CM onset even when a highly predictive variable, somatic cell count, was excluded from training and test data, but the models took longer to train. The effect of unmeasured variability on model performance demonstrates how predictors should be selected for RNNs, whereas RNNs appear to be very robust to misclassification in training data as well as missing variables. Researchers developing studies using deep learning should therefore focus their attention more on predictor selection than on reducing or standardizing outcome recording, since RNNs appear to be robust to the latter, while being more strongly impacted by the former.
The aim of the research. The domestic breeding base is not always able to meet the need for high-quality breeding stock with high genetic potential to staff newly established enterprises or farms that increase production capacity. Therefore, in recent decades, imports of livestock from abroad have increased significantly. Thus, according to the State Statistics Service of Ukraine, annually (2016–2021) from 1.5 to 4.5 thousand heads of breeding cattle are imported to our country from Europe (Denmark, the Netherlands, Germany, Poland, Austria, Hungary, the Czech Republic). The largest share of imported breeding resources are Holstein animals. However, imported animals that are obtained and raised in other environmental conditions and genetically programmed for them do not always successfully adapt to new housing and feeding conditions. It is obvious that the adaptation processes are reflected in the level of milk productivity and indicators of lifetime use. In this regard, further study of the productive longevity of cows and the realization of the genetic potential of imported and purchased within the country of dairy cattle is relevant and of scientific and practical interest. Materials and methods of research. The study was conducted in a breeding farm for breeding Ukrainian Black-and-White dairy cattle, and since 2009 – Holstein breed ALLC "Agrosvit" Myronivskyi district of Kyiv region by retrospective analysis on the materials of primary zootechnical and breeding records. The electronic information base of the Dairy Management System of Dairy Farming “ORSEC” as of March 2020 was used for the analysis. The generated matrix of observations in the sta format generally contained information about 5099 cows for 482 variables. Of these, 3298 animals had dated information on the date of calving (2002–2019) and milk yield of first heifer. Of the 1,001 cows included in the analysis, 541 were classified as Holstein, 541 as Ukrainian Black-and-White dairy cows, and 11 as other breeds and crossbreeds. By herd or country of selection 1135 cows are included in the group of local reproduction, 35 – imported in 2003 to the farm from Hungary (first calved in 2004), 105 cows imported in 2005 from Denmark (first calving 2005–2006), 33 cows imported in 2008 year from Germany (the first calving in 2008–2009), 48 cows were purchased from SE "Yamnytsia" Tysmenets district of Ivano-Frankivsk region, 20 – in SERF "Ryhalske" Yemilchyn district of Zhytomyr region, 53 – in LLC "Agrofirma Knyazhychi" Kyiv-Sviatoshynskyi district of Kyiv region, 33 – in Sarny SRS of Sarny district of Rivne region. Research results. Comparison of group average animals of different birthplaces (countries or herds of selection) established a sometimes noticeable level of intergroup differentiation in terms of growth intensity of repair heifers, reproductive ability and milk productivity of cows for the first three and higher lactation. This may be partly due to the different conditional bloodlines of the improving Holstein breed. In terms of live weight at the age of six months, the best development is characterized by animals of SE "Yamnytsia", which exceeded the animals of Sarny SRS by 18 ± 3.2 kg or 10.7% (td = 5.63, P < 0.001). The higher average group yield of first heifers is accompanied by a curvilinear increase in the duration of the service period and the period between calvings and a decrease in the coefficient of reproductive ability. The analysis of milk productivity for the second, third and higher lactation showed a significant advantage of imported animals from Germany over all other animals. Among the cows of domestic origin, the priority in milk yield for the second lactation are animals purchased from the SE "Yamnytsia". Among the animals of domestic selection, the most optimal indicators of lifetime use were characterized by cows purchased from SE "Yamnytsia", they prevailed in the number of lactations, duration of economic use, lactation, lifetime milk productivity (yield, fat, protein), milk yield and milk fat and protein per day of life, economic use and lactation not only all groups of animals born in Ukraine, but also cows of Danish and Hungarian selection. One-way analysis of variance confirmed the low, but in most cases significant influence of the place of birth of animals on the signs of duration and efficiency of lifetime use of cows. By age of the first calving, the difference between animals of European and domestic selection was insignificant (within the statistical error). In terms of efficiency of lifelong use, imported animals were not inferior to cows of Ukrainian selection, which showed a fairly high level of their adaptation to new economic and environmental conditions. A significant advantage of cows of European selection by the coefficient of economic (3.8 ± 0.99%, td = 3.84, P < 0.001) and productive (2.7 ± 0.88%, td = 3.07, < 0.001) use at a lower lactation factor (1.7 ± 0.84%, td = 2.02, P < 0.05). Conclusions. According to the intensity of growth of repair heifers, reproductive capacity and milk productivity of cows for the first three and higher lactation, sometimes a noticeable level of intergroup differentiation of animals of different birthplaces (countries or herds of selection) was established. In terms of live weight of heifers, animals of Hungarian selection, local reproduction and purchased from SE “Yamnytsia” had the advantage, the worst were peers from Sarny SRS. The youngest age of calving was characterized by the first heifers of German selection and local reproduction (ALLC "Agrosvit"). The highest milk productivity for the first lactation was distinguished by cows of Hungarian, for the second and older – of German selection, local reproduction and purchased from SE "Yamnytsia". The tendency of deterioration of reproductive ability of cows with increase of their dairy productivity is revealed.
The influence of genotypic factors (breed, Holstein share heredity, line or related group, sire) on the longevity and lifetime production of dairy cows has been investigated. The study was carried out in a retrospective statistical experiment on the commercial dairy farm “Terezyne”, which located in Kyiv region. The formed matrix of observations in the sta format generally contained information about 5703 cows for 458 variables. Holstein cows were characterized by higher longevity (1489 ± 27.8 days) and higher lifetime milk production (21940 ± 500.9 kg), the lowest longevity and lifetime milk production had cows of Ukrainian Red-and-White dairy breed. Holstein cows compared to animals of Ukrainian Red-and-White dairy breed had in average more lactations (+0.57 ± 0.187) and higher number of total calves (+0.59 ± 0.244). They had an advantage in lifespan by 218 ± 80.1 days, productive lifespan – by 326 ± 82.0 days, total lactation length – by 282 ± 62.2 days. During the lifetime, they produced 4119 ± 1398.4 kg more milk and 549.3 ± 99.62 kg more milk fat and milk protein. Lifetime daily milk yield of Holstein cows was on average 2.6 ± 0.37 kg higher, lifetime daily milk fat and milk protein – 208 ± 26.7 g more compared to animals of the Ukrainian Red-and-White dairy breed. Cows of Ukrainian Black-and-White dairy breed in most traits were intermediate between Holstein and Ukrainian Red-and-White dairy breed. While increasing Holstein share heredity the longevity and lifetime production of cows tended for curvilinear growing. Lifetime of Holstein cows (100%) in the herd was 0.16 ± 0.405 lactations and 325 ± 170.7 days (P < 0.1) longer compared to animals with Holstein share heredity less than 75%. Their lifetime milk production was 8969 ± 2351.6 kg (P < 0.001) higher, lifetime milk fat and milk protein production by 709.9 ± 164.73 kg (P < 0.001) higher, lifetime daily yield by 3.0 ± 0.62 kg (P < 0.001) and 250 ± 43.5 g (P < 0.001), respectively, higher. The cows of Starbuck 352790, Valiant 1650414 and Elevation 1491007 lines and daughter of V. Astronomer 2160438 and H. R. Artist 6284191 sires showed the best longevity and lifetime production. The advantage of cows of the related group of Starbuck 352790 in comparison to animals of other lines in total number of lactations was 0.28–1.29, number of total calves – 0.16–1.20, lifespan – 90–508 days, productive lifespan – 116–603 days, total lactation length – 98–500 days, lifetime milk production – 1402–12161 kg, lifetime milk fat and milk protein production – 115–892 kg. Daughters of bull V. Astronomer 2160438 characterized by 0.96–2.72 more total number of lactations than daughters of other bulls; they had 0.80–2.33 higher average number of total calves. In terms of lifespan, productive lifespan and total lactation length, daughters of this bull were predominated the daughters of other bulls by 392–1037, 297–1143 and 278–971 days, respectively. One-way analysis of variance showed that the greatest influence on longevity and lifetime production of dairy cows had a paternal inheritance, the lowest – breed, it means there is a tendency of increasing the influence of genotypic factors on the researched traits while reducing the level of selection group in the system hierarchy. The strength of breed influence (higher level of intraspecific selection system hierarchy) on the traits of longevity and lifetime production of cows ranged from 1.1 to 12.5%, Holstein share heredity – 3.9 to 19.5%, line or related group – 4.0–19.8% and paternal inheritance – 25.0–47.6%. Greater degree of influence of paternal inheritance is explained both by the lowest (basic) level of intrabreed system hierarchy (the closest level of intragroup kinship) and (partially) by a much higher number of gradations of the organized factor.
Hypocalcemia in parturient dairy cows causes progressive neuromuscular dysfunction with flaccid paralysis, circulatory collapse, and depression of consciousness. Hypocalcemia in goats causes combinations of tetany and flaccid paralysis. Clinical cases of milk fever generally respond well to treatment; however, costs are high owing to lingering complications. Subclinical hypocalcemia is costlier because it affects a high proportion of older animals. The pathogenesis of hypocalcemia is complex and involves several interdependent pathways. Strategies for preventing hypocalcemia include dietary calcium restriction (or feeding calcium binders), dietary acidification, limiting dietary phosphorus, increasing dietary magnesium or sulfur, vitamin D supplementation, and prophylactic calcium administration.
Stan is a probabilistic programming language for specifying statistical models. A Stan program imperatively defines a log probability function over parameters conditioned on specified data and constants. As of version 2.14.0, Stan provides full Bayesian inference for continuous-variable models through Markov chain Monte Carlo methods such as the No-U-Turn sampler, an adaptive form of Hamiltonian Monte Carlo sampling. Penalized maximum likelihood estimates are calculated using optimization methods such as the limited memory Broyden-Fletcher-Goldfarb-Shanno algorithm. Stan is also a platform for computing log densities and their gradients and Hessians, which can be used in alternative algorithms such as variational Bayes, expectation propagation, and marginal inference using approximate integration. To this end, Stan is set up so that the densities, gradients, and Hessians, along with intermediate quantities of the algorithm such as acceptance probabilities, are easily accessible. Stan can be called from the command line using the cmdstan package, through R using the rstan package, and through Python using the pystan package. All three interfaces support sampling and optimization-based inference with diagnostics and posterior analysis. rstan and pystan also provide access to log probabilities, gradients, Hessians, parameter transforms, and specialized plotting.
This paper discusses statistical modelling for data with a hierarchical structure, and distinguishes in this context between three different meanings of the term hierarchical model: to account for clustering, to investigate variability and separate predictive equations at different hierarchical levels (multi-level analysis), and in a Bayesian framework to involve multiple layers of data or prior information. Within each of these areas, the paper reviews both past developments and the present state, and offers indications of future directions. In a worked example, previously reported data on piglet lameness are reanalyzed with multi-level methodology for survival analysis, leading to new insights into the data structure and predictor effects. In our view, hierarchical models of all three types discussed have much to offer for data analysis in veterinary epidemiology and other disciplines.
Risk of culling consequent to the main health disorders occurring in the current production systems is reviewed. Survival analysis including health disorders as time-dependent variables is considered to be the most appropriate method to assess their effects because they allow a better description of the exact follow-up of disease history. Farmers preferentially consider health events in the current lactation and/or those occurring in early stages of lactation for making culling decisions. The unfavourable direct effects on culling of dystocia and udder disorders (mastitis and teat injuries) are clearly demonstrated, whereas there are variations between studies on the association between metabolic and reproductive disorders and culling. These variations may be due to differences in study designs, populations involved and methods. Consequences, in terms of estimated effect of health disorders, of methodological choices (e.g. whether or not including in the models descriptors for milk yield and/or reproductive performance) are discussed. Metabolic and reproductive disorders may act indirectly through a subsequent decrease in milk yield and reproductive performance. The impact of health disorders on longevity is on average weak, compared to the impact of low milk yield potential and poor reproductive performance. Herd characteristics (availability of heifers, quota, farmer's attitude towards risk and uncertainty...) modify the risk for a cow to be culled for a given health disorder. Aims of further studies could be (1) to interpret the meaning and to analyse the reliability of culling reasons information, (2) to evaluate the relative effect on culling of health disorders and performance (milk yield and reproduction) in different parities, (3) to investigate the role of components of the herd effect on the risk of culling.
Knowledge of reproductive risk factors for culling is useful in making insemination and culling decisions and helps motivate efforts to reduce or eliminate risk factors. The objective of this study was to describe survival and reproductive risk factors for culling in Holstein dairy herds with at least 200 cows. Results were calculated from 2,345,015 DHI lactation records from 727 herds with at least 200 cows from 2001 to 2006. Herds were located in 36 states primarily located east of the Mississippi River. Kaplan-Meier survival curves were obtained and daily hazards of culling were calculated with the actuarial method. Cox regression was performed with the GLIMMIX procedure in SAS (SAS Institute Inc., Cary, NC). The hazard of culling increased with parity number. Cows in their sixth parity had 3 times greater hazards than cows in their first parity. Medium remaining productive life for cows calving in parity 1 to 6 were 907, 697, 553, 469, 423, and 399 d, respectively. Daily hazards of culling first peaked approximately 30 d after calving and then again later in lactation, after 280 d, for older cows. Hazards for first-parity cows peaked earlier, around d 10 after calving, and the first-parity cows had lower risks of culling later in lactation than older cows. Pregnant cows had 3 to 7 times lower hazards of culling than open cows. Hazards of culling increased for cows that had greater calving difficulty, gave birth to males or twins, were in herds with shorter days to first insemination, or had longer days to conception. The possible to likely use of a synchronized breeding program increased from 21.9% in 2001 to 41.4% in 2006. Cows in herds that did not use a synchronized breeding program had slightly lower risks of culling than those in herds that at least possibly used a synchronized breeding program.
The objective of most biomedical research is to determine an unbiased estimate of effect for an exposure on an outcome, i.e. to make causal inferences about the exposure. Recent developments in epidemiology have shown that traditional methods of identifying confounding and adjusting for confounding may be inadequate.
The traditional methods of adjusting for "potential confounders" may introduce conditional associations and bias rather than minimize it. Although previous published articles have discussed the role of the causal directed acyclic graph approach (DAGs) with respect to confounding, many clinical problems require complicated DAGs and therefore investigators may continue to use traditional practices because they do not have the tools necessary to properly use the DAG approach. The purpose of this manuscript is to demonstrate a simple 6-step approach to the use of DAGs, and also to explain why the method works from a conceptual point of view.
Using the simple 6-step DAG approach to confounding and selection bias discussed is likely to reduce the degree of bias for the effect estimate in the chosen statistical model.
A survey was carried out in a random sample of 123 dairy farms from the east of Ireland. The monthly mean production per cow was 315 l of milk and 11.5 kg of fat. The mean log herd somatic cell count was 5.45 (arithmetic mean = 372,573 cells/ml), with almost 50% of the monthly counts over 300,000 cells/ml in a 12-month period. Bivariate and multivariate analysis was performed to assess the relative impact of the personal characteristics of the farmer and the management policies he applied on the amount and quality of the milk produced. In five out of six models the group of variables related to farmers' attitudes, values, and sociodemographic profile explained a similar or greater amount (between 14.44 and 34.35%) of the variation of farm performance than the group of management variables (between 14.33 and 25.99%) as measured by the R2. These results stress the importance of the human factors in explaining variation in farm performance.
Data from a survey performed from 1986 to 1990 were analyzed to assess the effects of diseases on length of productive life of 3589 Holstein cows from 47 herds, using a proportional hazard model. The probability of a cow being culled, or hazard function, was supposed to be the product of an unspecified baseline hazard function and log-linear, time-dependent explanatory variables that possibly influence culling rate (Cox's regression). The effect of 16 health events was studied according to lactation number of occurrence. The model included adjustments for effects of herd-year-season (considered to be random), month of calving, stage of lactation, lactation number, reproductive performance, and milk production. The probability of a cow being culled increased in early and late stages of lactation in older cows, in low producing cows, and in cows with poor reproductive performance. Mastitis before the peak of lactation or during the dry period increased the risk (relative culling rate in first lactation, 1.3 and 4.0, respectively). Teat injuries and nontraumatic udder disorders had a large impact on longevity. Cows with late metritis or early abortion had poor survival. The decrease in median length of productive life could be over a standard lactation in particular cases. Expected survivor curves, computed after assumption of a priori values of covariates and their evolution over time, appear to be powerful tools for examining the effect of health disorders on length of productive life of cows.
The effect of seven diseases on culling was measured in 7523 Holstein cows in New York State. The cows were from 14 herds and had calved between January 1, 1994 and December 31, 1994; all cows were followed until September 30, 1995. Survival analysis was performed using the Cox proportional hazards model to incorporate time-dependent covariates for diseases. Different intervals representing stages of lactation were considered for effects of the diseases. Five models were fitted to test how milk yield and conception status modified the effect of diseases on culling. Covariates in the models included parity, calving season, and time-dependent covariates measuring diseases, milk yield of the current lactation, and conception status. Data were stratified by herd. The seven diseases and lactational risks under consideration were milk fever (0.9%), retained placenta (9.5%), displaced abomasum (5.3%), ketosis (5.0%), metritis (4.2%), ovarian cysts (10.6%), and mastitis (14.5%). Older cows were at a much higher risk of being culled. Calving season had no effect on culling. Higher milk yield was protective against culling. Once a cow had conceived again, her risk of culling dropped sharply. In all models, mastitis was an important risk factor throughout lactation. Milk fever, retained placenta, displaced abomasum, ketosis, and ovarian cysts also significantly affected culling at different stages of lactation. Metritis had no effect on culling. The magnitude of the effects of the diseases decreased, but remained important, when milk yield and conception status were included as covariates. These results indicated that diseases have an important impact on the actual decision to cull and the timing of culling. Parity, milk yield, and conception status are also important factors in culling decisions.
The objectives of this presentation are to review results of our previous and on-going research with respect to the risk factors and consequences of poor reproductive performance in dairy cows, and to develop an economic framework to optimize decisions related to dairy cow reproductive performance. To make profitable breeding and replacement decisions, the farmer must account for factors including age, production level, lactation stage, pregnancy status, and disease history of the cows in the herd. Establishing the interrelationships among disease, milk yield, reproduction, and herd management is necessary for developing a decision model for disease treatment, insemination, and replacement.
Dairy Herd Improvement Holstein herd summary records (n = 11,259) were obtained for the year ending 1998. Reasons cows reportedly left the herd based on termination codes were analyzed for the effect of region, herd size, and herd milk production level. Regions were: North, Midsouth, and South. Herd sizes were: small (25 to 99), low medium (100 to 149), high medium (150 to 299), and large (greater than or equal to 300 cows). Milk production levels were: low (less than 7258 kg), medium (7258 to 9072 kg), and high (greater than 9072 kg). The overall percentage of cows leaving the herd was higher in the Midsouth than the South and increased with herd size. Low producing herds reported a lower percentage of cows left than high producing herds. Herds in the South reported more cows leaving for reproduction, death, and low production and fewer leaving for mastitis. Herds in the North and Midsouth reported more cows leaving for injury/other and disease, respectively. Cows left herds for disease less frequently in the North. Large herds in the South had a higher percentage leaving for low production than any herd size group in the North. Small herds reported more cows leaving for reproduction and mastitis than high medium and low medium size herds. The percentage of cows leaving for feet and leg problems was lowest for small size herds. High producing herds reported more cows leaving for reproduction, mastitis, feet and legs and disease.
This paper describes the core features of the R package geepack, which implements the generalized estimating equations (GEE) approach for fitting marginal generalized linear models to clustered data. Clustered data arise in many applications such as longitudinal data and repeated measures. The GEE approach focuses on models for the mean of the correlated observations within clusters without fully specifying the joint distribution of the observations. It has been widely used in statistical practice. This paper illustrates the application of the GEE approach with geepack through an example of clustered binary
data.
The relationship between cows’ health, reproductive performance or disorders and their longevity is well demonstrated in the literature. However these associations at the cow level might not hold true at the herd level, and herd-level variables can modify cow-level outcomes independently of the cows’ characteristics. The interaction between cow-level and herd-level variables is a relevant issue for understanding the culling of dairy cows. However it requires the appropriate group-level variables to assess any contextual effect. Based on 10 years of health and production data, the objectives of this paper are:(a) to quantify the culling rates of dairy herds in Québec; (b) to determine the profiles of the herds based on herd-level factors, such as demographics, reproduction, production and health indicators, and whether these profiles can be related to herd culling rates for use as potential contextual variables in multilevel modelling of culling risk. A retrospective longitudinal study was conducted on data from dairy herds in Québec, Canada, by extracting health information events from the dairy herd health management software used by most Québec producers and their veterinarians. Data were extracted for all lactations taking place between January 1st, 2001 and December 31st, 2010. A total of 432,733 lactations from 156,409 cows out of 763 herds were available for analysis. Thirty cow-level variables were aggregated for each herd and years of follow-up, and their relationship was investigated by Multiple Factor Analysis (MFA). The overall annual culling rate was 32%, with a 95% confidence interval (CI) of [31.6%,32.5%]. The dairy sale rate by 60 days in milk (DIM) was 3.2% [2.8%,3.6%]. The annual culling rate within 60 DIM was 8.2% [7.9%,8.4%]. The explained variance for each axis from the MFA was very low: 14.8% for the first axis and 13.1% for the second. From the MFA results, we conclude there is no relationship between the groups of herd-level indicators, demonstrating the heterogeneity among herds for their demographics, reproduction and production performance, and health status. However, based on Principal Component Analysis (PCA), the profiles of herds could be determined according to specific, single, herd-level indicators independently. The relationships between culling rates and specific herd-level variables within factors were limited to livestock sales, proportion of first lactation cows, herd size, proportion of calvings occurring in the fall, longer calving intervals and reduced 21-day pregnancy rates, increased days to first service, average age at first calving, and reduced milk fever incidence. The indicators found could be considered as contextual variables in multilevel model-building strategies to investigate cow culling risk.
Evaluation of herd culling and replacement strategies is an element of profitable dairy herd production medicine programs. A wealth of scientific literature unanimously concludes that, for many herds, standard North American dairy industry culling practices may maximize production but not profit. To ensure profitable client recommendations, dairy practitioners need to evaluate their assumptions and beliefs about culling and replacement strategies and understand the economic principles underlying these strategies.
An assessment of the importance of farm managers' attitudes, i.e., the farm managers' socio-psychological characteristics, relative to management practices, and with respect to farm performance on southwestern Ontario dairy farms is presented. The assessment was done using a multiple linear regression analysis. Eight dependent variables measuring farm performance were used; namely, retained placenta (%), metritis (%), ovarian disorders (%), other reproductive disorders (%), calving interval (months), culling (%) and herd BCA (breed class average) for fat and milk. Socio-psychological variables appeared in the best regression models of each of the dependent variables, and were found to be quantitatively as important or more so than the management group of variables. In the best regression model for the first six dependent variables, the group of socio-psychological variables explained between 10.8% and 24.5% of the variation in the dependent variable as compared with 0% to 15.5% for the management group of variables. The best regression model for the last two dependent variables showed a synergistic association between the two groups of independent variables.
The results of this study indicate that farm managers' attitudes are important when studying farm performance. Attitudes should be measured and considered before proposing management practices to improve farm performance since interactions between attitudes and management practices suggest that attitudes act as effect modifers on the management practices-herd performance relationship. However, further research is needed to better understand the mode of action of managers' attitudes in the dairy farm system.
In multilevel modeling the residual variation in a response variable is split into component parts that are attributed to various levels. In applied work, much use is made of the percentage of variation that is attributable to the higher level sources of variation. Such a measure, however, makes sense only in simple variance components, Normal response, models where it is often referred to as the intra-unit correlation. In this article we describe how similar measures can be found for both more complex random variation in Normal response models and models with discrete responses. In these cases the variance partitions are dependent on predictors associated with the individual observation. We compare several computational techniques to compute the variance partitions.
Hamiltonian dynamics can be used to produce distant proposals for the
Metropolis algorithm, thereby avoiding the slow exploration of the state space
that results from the diffusive behaviour of simple random-walk proposals.
Though originating in physics, Hamiltonian dynamics can be applied to most
problems with continuous state spaces by simply introducing fictitious
"momentum" variables. A key to its usefulness is that Hamiltonian dynamics
preserves volume, and its trajectories can thus be used to define complex
mappings without the need to account for a hard-to-compute Jacobian factor - a
property that can be exactly maintained even when the dynamics is approximated
by discretizing time. In this review, I discuss theoretical and practical
aspects of Hamiltonian Monte Carlo, and present some of its variations,
including using windows of states for deciding on acceptance or rejection,
computing trajectories using fast approximations, tempering during the course
of a trajectory to handle isolated modes, and short-cut methods that prevent
useless trajectories from taking much computation time.
The occurrence of five periparturient events and their effects on subsequent culling and fertility was recorded in eight herds in the UK. Combining data from all 2105 calvings, the proportion affected by assisted calving, dead calf, retained fetal membranes (RFM), milk fever or twins was 5.9, 8.2, 5.3, 5.0 or 3.3 per cent, respectively. Compared with unaffected herdmates, cows with an assisted calving or a dead calf had higher early (but not late) culling rates, (assisted calving: 8.8 per cent being culled before 100 days after calving compared with 5.7 per cent; P=0.05; dead calf: 12.2 per cent culled compared with 5.3 per cent; P=0.001). Compared with unaffected animals, cows with milk fever were four times more likely to be culled before 100 days after calving (16.2 per cent compared with 5.3 per cent; P=0.001), whereas those with RFM were twice as likely to be culled between 100 and 200 days (14.3 per cent compared with 7.6 per cent; P=0.003), and both groups were twice as likely to not be pregnant by 200 days. Cows with RFM or milk fever also had markedly reduced subsequent fertility: both conditions extended calving to pregnancy intervals (by 20 days; P=0.001, or by 13 days; P=0.03, respectively), lowered 100-day in-calf rates (by 24.5 per cent; P=0.001, or by 17.8 per cent; P=0.008, respectively) and lowered 200-day in-calf rates (by 20 per cent; P=0.001, or by 15 per cent; P=0.002, respectively). The birth of twins had no effect on subsequent culling or fertility.
Data from an observational study, carried out during a 4.5 year period (1986–1990), were used to quantify the effects of health disorders on the risk of culling. The study population consisted of 47 commercial Holstein dairy herds from western France, comprising 4123 cows.Logistic regression was used to assess the relationships between health disorders and early and late culling. Fourteen main health disorders with clinical signs and one subclinical disease were studied: abortion, periparturient accident, calving provided with assistance, digestive disorders, ketosis, locomotor disorders, mastitis, metritis, milk fever, cystic ovaries, respiratory disorders, retained placenta, teat injuries, non-traumatic udder disorders and status with respect to milk somatic cell count. Adjustments were made for year, month of calving, parity, breeding value for milk, best of the two first milk production records and reproductive performance. The possible effects of interactions among variables were also studied. The herd effect was taken into account using random effect models.Non-traumatic udder disorders, teat injuries, milk fever and the occurrence of both ketosis and assistance at calving were significantly associated with an increased risk of being early culled (odds ratios (OR) ranging from 1.6 to 10.3). Early and late abortion, late metritis, poor peproductive performance, retained placenta, non-traumatic udder disorders within 45 days post-partum and mastitis occurring in the first 3 months of the lactation were positively associated with a late culling (OR ranging from 1.2 to 6.6). Cows with lower breeding value for milk and higher parities were high risk groups for culling. A lower level of milk production and occurrence of both reproductive disorders and poor reproductive performance were risk factors for late culling.
Hamiltonian Monte Carlo (HMC) is a Markov chain Monte Carlo (MCMC) algorithm
that avoids the random walk behavior and sensitivity to correlated parameters
that plague many MCMC methods by taking a series of steps informed by
first-order gradient information. These features allow it to converge to
high-dimensional target distributions much more quickly than simpler methods
such as random walk Metropolis or Gibbs sampling. However, HMC's performance is
highly sensitive to two user-specified parameters: a step size {\epsilon} and a
desired number of steps L. In particular, if L is too small then the algorithm
exhibits undesirable random walk behavior, while if L is too large the
algorithm wastes computation. We introduce the No-U-Turn Sampler (NUTS), an
extension to HMC that eliminates the need to set a number of steps L. NUTS uses
a recursive algorithm to build a set of likely candidate points that spans a
wide swath of the target distribution, stopping automatically when it starts to
double back and retrace its steps. Empirically, NUTS perform at least as
efficiently as and sometimes more efficiently than a well tuned standard HMC
method, without requiring user intervention or costly tuning runs. We also
derive a method for adapting the step size parameter {\epsilon} on the fly
based on primal-dual averaging. NUTS can thus be used with no hand-tuning at
all. NUTS is also suitable for applications such as BUGS-style automatic
inference engines that require efficient "turnkey" sampling algorithms.
The objective was to quantify the effect of postpartum uterine diseases on milk production and culling. Data from 2,178 Holstein cows in 6 herds enrolled in a randomized clinical trial were used. Milk production data from the first 4 Dairy Herd Improvement Association (DHIA) test-days and culling data from farm records were collected. Retained placenta (RP; ≥24 h after parturition) and metritis [≤20 d in milk (DIM)] were diagnosed by farm managers using standardized definitions. Farms were visited weekly and cows were examined at 35 and 56 (±3) DIM using endometrial cytology (cytobrush device), vaginal discharge scoring (Metricheck device), and measurement of cervical diameter by transrectal palpation. Diagnostic criteria for cytological endometritis (CYTO) and purulent vaginal discharge (PVD) were established based on a detrimental effect on subsequent reproduction. Statistical analyses were performed using linear mixed models, logistic regression models, and Cox proportional hazard models, accounting for the effects of experimental treatments and herd clustering. Milk production and culling were the outcomes. Primiparous and multiparous cows were modeled separately for milk production. Milk production of primiparous cows was unaffected by uterine diseases. The effect of metritis on milk production was variable over time in multiparous cows: it decreased production per cow by 3.7 kg at the first DHIA test, but was not different at later tests. Retained placenta decreased milk production by 2.6 kg/d in multiparous cows through the first 4 DHIA tests. The projected effects of metritis and RP in multiparous cows were reductions of 259 kg and 753 kg over 305 DIM, respectively; these effects were additive. Neither CYTO nor PVD affected milk production. Culling risks at 30 and 63 DIM were unaffected by RP and metritis. Culling hazard up to 300 DIM was unaffected by RP, metritis, CYTO, or PVD, whether or not pregnancy status, milk production, and displaced abomasum were accounted for. Uterine disease decreased pregnancy rate, which was a substantial risk factor for culling; however, if affected cows became pregnant they were not at greater risk of culling.
The objective of this observational study was to investigate the risk factors for metritis, purulent vaginal discharge, and cytological endometritis. The hypothesis was that purulent vaginal discharge and cytological endometritis would have different risk factors because they represent distinct manifestations of uterine disease. Data generated from 1,363 Holstein cows (3 herds) enrolled in a randomized clinical trial were used. Calving history, periparturient disease incidence, and body condition score at calving and at 63 d in milk (DIM) were recorded. Serum nonesterified fatty acid concentration was measured once during the week before expected calving. Serum nonesterified fatty acid, β-hydroxybutyric acid, and haptoglobin (Hapto) concentrations were measured at 4 ± 3, 11 ± 3, and 18 ± 3 DIM. Serum progesterone concentration was measured at 21 ± 3, 35 ± 3, 49 ± 3, and 63 ± 3 DIM. Metritis was diagnosed by farm managers within the first 20 DIM using a standardized definition. Cows were examined at 35 ± 3 DIM by a veterinarian for purulent vaginal discharge (mucopurulent or worse vaginal discharge; Metricheck device) and cytological endometritis (≥ 6% polymorphonuclear cells on endometrial cytology; cytobrush device). Statistical analyses were performed using multivariable logistic regression models for each disease, accounting for the random effect of herd. Risk factors for metritis included increased nonesterified fatty acid prepartum (≥ 0.6 mmol/L), dystocia, retained placenta, and increased Hapto in the first week postpartum (≥ 0.8 g/L). Risk factors for purulent vaginal discharge included twinning, dystocia, metritis, and increased Hapto (≥ 0.8 g/L) in the first week postpartum. Risk factors for cytological endometritis included low body condition score at parturition (≤ 2.75), hyperketonemia (≥ 1,100 μmol/L), and increased Hapto (≥ 0.8 g/L) in the first week postpartum. These results support the hypothesis that some of the risk factors for purulent vaginal discharge and cytological endometritis are different, which supports that they are distinct manifestations of uterine disease.
In regular statistical models, the leave-one-out cross-validation is asymptotically equivalent to the Akaike information criterion. However, since many learning machines are singular statistical models, the asymptotic behavior of the cross-validation remains unknown. In previous studies, we established the singular learning theory and proposed a widely applicable information criterion, the expectation value of which is asymptotically equal to the average Bayes generalization loss. In the present paper, we theoretically compare the Bayes cross-validation loss and the widely applicable information criterion and prove two theorems. First, the Bayes cross-validation loss is asymptotically equivalent to the widely applicable information criterion as a random variable. Therefore, model selection and hyperparameter optimization using these two values are asymptotically equivalent. Second, the sum of the Bayes generalization error and the Bayes cross-validation error is asymptotically equal to $2\lambda/n$, where $\lambda$ is the real log canonical threshold and $n$ is the number of training samples. Therefore the relation between the cross-validation error and the generalization error is determined by the algebraic geometrical structure of a learning machine. We also clarify that the deviance information criteria are different from the Bayes cross-validation and the widely applicable information criterion.
A large portion of current epidemiologic research is based on methodologic individualism: the notion that the distribution of health and disease in populations can be explained exclusively in terms of the characteristics of individuals. The present paper discusses the need to include group- or macro-level variables in epidemiologic studies, thus incorporating multiple levels of determination in the study of health outcomes. These types of analyses, which have been called contextual or multi-level analyses, challenge epidemiologists to develop theoretical models of disease causation that extend across levels and explain how group-level and individual-level variables interact in shaping health and disease. They also raise a series of methodological issues, including the need to select the appropriate contextual unit and contextual variables, to correctly specify the individual-level model, and, in some cases, to account for residual correlation between individuals within contexts. Despite its complexities, multilevel analysis holds potential for reemphasizing the role of macro-level variables in shaping health and disease in populations.
The report upon which the current discussion is based was prepared in response to the increasing interest of the dairy industry in the recording of clinical disease data. The major objective was to introduce guidelines and standards for the recording and presentation of the diseases of dairy cattle. Eight clinically identifiable diseases of economic importance to the dairy industry were considered: milk fever, retained placenta, metritis, ketosis, left displaced abomasum, cystic ovarian disease, lameness, and clinical mastitis. Standardized definitions for these diseases were established through consultation with industry partners. Two approaches to summarization and reporting were proposed. For retrospective analysis, which is used when historical data are summarized for genetic evaluation for example, lactational incidence risk (cumulative incidence) has been recommended. For current analysis, which is used for herd health monitoring, a true incidence rate has been recommended. Milk fever and retained placenta were exceptions to the latter because of their short periods of risk. For these two diseases, lactational incidence risks are reported.
Incidences of diseases and their effects on reproductive performance and risk of culling in herds stratified by production and estrus detection efficiency were studied. Data were from the Swedish milk and disease recording systems and consisted of records for 33,748 first parity Swedish Friesian cows. A standardized mixed threshold model was used for statistical analyses of categorical outcome variables, and an ordinary linear mixed model was used for continuous outcome variables. An increase in production was associated with increased frequencies of treatments of most diseases, shorter intervals from calving to first artificial insemination, fewer days open, and lower culling rates. Cows treated for metritis, silent estrus, and cystic ovaries had an increased number of days to first artificial insemination and more days open. However, the negative consequences of these diseases on reproductive performance decreased as herd production increased. The risk of culling was higher for cows treated for dystocia, cystic ovaries, and mastitis, but the increase in the risk of culling was lower for higher producing herds. Similar trends were observed when herds were stratified by estrus detection efficiency. The results support the hypothesis that herd management, as characterized by milk production or estrus detection efficiency, is important in the incidences and consequences of diseases. Herd management, measured directly or indirectly, should be considered when the health status or cost of disease for a given herd is evaluated.
Causal diagrams have a long history of informal use and, more recently, have undergone formal development for applications in expert systems and robotics. We provide an introduction to these developments and their use in epidemiologic research. Causal diagrams can provide a starting point for identifying variables that must be measured and controlled to obtain unconfounded effect estimates. They also provide a method for critical evaluation of traditional epidemiologic criteria for confounding. In particular, they reveal certain heretofore unnoticed shortcomings of those criteria when used in considering multiple potential confounders. We show how to modify the traditional criteria to correct those shortcomings.
The effects of 15 diseases on time until culling were studied in 39,727 Finnish Ayrshire cows that calved during 1993 and were followed until the next calving or culling. The diseases studied were: dystocia, milk fever, retained placenta, displacement of the abomasum, metritis, non-parturient paresis, ketosis, rumen disorders, acute mastitis, hypomagnesemia, lameness, traumatic reticuloperitonitis, anestrus, ovarian cysts, and teat injuries. Survival analysis, using the Cox proportional hazards model, was performed and diseases were modeled as time-dependent covariates. Different stages of lactation when culling can occur were also considered. Parity, calving season and herd were included as covariates in every model. Parity had a significant effect on culling, the risk of culling being four times higher for a cow in her sixth or higher parity than for a first parity cow. The effects of diseases varied according to when the diseases occurred and when culling occurred. Mastitis, teat injuries and lameness had a significant effect on culling throughout the whole lactation. Anestrus and ovarian cysts had a protective effect against culling at the time when they were diagnosed. In general, diseases affected culling decisions mostly at the time of their occurrence. The effect seemed to decrease with time from the diagnosis of the disease. However, milk fever, dystocia and metritis also had a significant effect on culling at the end of the lactation.
Effects on reproduction of dystocia, stillbirth, abortion, milk fever, retained placenta, metritis, cystic ovaries, anestrus, ketosis, displaced abomasum, locomotor disorders, and mastitis were reviewed. Papers were considered if they provided quantitative estimates of diseases on days to first estrus, days to first service, conception rate at first service, days from first service to conception, days to conception or days open, calving interval, conception rates at various days post partum (dpp), and number of services per conception or per cow. Only papers in English in peer-reviewed journals were selected for analysis of post 1960 data from intensive dairy regions. Seventy papers fulfilled the selection criteria. Summary estimates of disease effects were calculated according to meta-analysis methods, and study designs were described in detail to identify possible heterogeneity of the results. Stillbirth, milk fever, displaced abomasum and mastitis had no effect on reproduction. Clinical ketosis, dystocia and retained placenta were associated with 2 to 3 more days to first service and with a 4 to 10% lower conception rate at first service, resulting in 6 to 12 more days to conception. Locomotor disorders were associated with an average increase of 12 d to conception, with wide variation depending on lesions and stage of occurrence. Metritis was associated with 7 more days to first service, 20% lower conception rate at first service, resulting in 19 more days to conception. Cystic ovaries were associated with 6 to 11 more days to first service and with 20 to 30 more days to conception. Anestrus was associated with 26 more days to first service and with an 18% lower conception rate at first service, resulting in 41 more days to conception. Abortion was associated with 70 to 80 more days to conception.
Logistic regression with random effects is used to study the relationship between explanatory variables and a binary outcome in cases with nonindependent outcomes. In this paper, we examine in detail the interpretation of both fixed effects and random effects parameters. As heterogeneity measures, the random effects parameters included in the model are not easily interpreted. We discuss different alternative measures of heterogeneity and suggest using a median odds ratio measure that is a function of the original random effects parameters. The measure allows a simple interpretation, in terms of well-known odds ratios, that greatly facilitates communication between the data analyst and the subject-matter researcher. Three examples from different subject areas, mainly taken from our own experience, serve to motivate and illustrate different aspects of parameter interpretation in these models.
Logistic regression models were used to examine the relationship between milk yield and incidence of certain disorders. Lactations (n = 2197) of 1074 Holstein-Friesian cows from 10 dairies (25 to 146 cows per dairy) in Lower Saxony were studied. The 305-d yield from the previous and current lactations served as the standards for milk yield. Eight disorder complexes were considered: retained placenta, metritis, ovarian cysts, mastitis, claw diseases, milk fever, ketosis, and displaced abomasum. Each disorder complex was modeled separately. In addition to milk yield, the influences of the lactation number, the calving season and the other disorder complexes were examined with the "herd" factor taken into account. A correlation between retained placenta, mastitis, and milk fever to milk yield during the previous lactation was found to be probable and for ketosis and displaced abomasum such a correlation was found to be possible. A connection to the yield in the current lactation was shown for ovarian cysts, claw diseases, and milk fever. No relationship to milk yield existed for metritis. An influence of the lactation number was also demonstrated in various models. Single models allowed a demonstration of the influences of both milk yield and lactation number. Limitations of the model types are discussed.
This paper investigates generalized estimating equations for association parameters, which are frequently of interest in family studies, with emphasis on covariance estimation. Separate link functions are used to connect the mean, the scale, and the correlation to linear predictors involving possibly different sets of covariates, and separate estimating equations are proposed for the three sets of parameters. Simulations show that the robust 'sandwich' variance estimator and the jackknife variance estimator for the correlation parameters are generally close to the empirical variance for the sample size of 50 clusters. The results contradict Ziegler et al. and Kastner and Ziegler, where the 'sandwich' estimator obtained from the software MAREG was shown to be unsuitable for practical usage. The problem appears to arise because the MAREG variance estimator does not account for variability in estimation of the scale parameters, but may be valid with fixed scale. We also find that the formula for the approximate jackknife variance estimator in Ziegler et al. is deficient, resulting in systematic deviations from the fully iterated jackknife variance estimator. A general jackknife formula is provided and performs well in numerical studies. Data from a study on the genetics of alcoholism is used to illustrate the importance of reliable variance estimation in biomedical applications.
The logistic regression model is frequently used in epidemiologic studies, yielding odds ratio or relative risk interpretations. Inspired by the theory of linear normal models, the logistic regression model has been extended to allow for correlated responses by introducing random effects. However, the model does not inherit the interpretational features of the normal model. In this paper, the authors argue that the existing measures are unsatisfactory (and some of them are even improper) when quantifying results from multilevel logistic regression analyses. The authors suggest a measure of heterogeneity, the median odds ratio, that quantifies cluster heterogeneity and facilitates a direct comparison between covariate effects and the magnitude of heterogeneity in terms of well-known odds ratios. Quantifying cluster-level covariates in a meaningful way is a challenge in multilevel logistic regression. For this purpose, the authors propose an odds ratio measure, the interval odds ratio, that takes these difficulties into account. The authors demonstrate the two measures by investigating heterogeneity between neighborhoods and effects of neighborhood-level covariates in two examples--public physician visits and ischemic heart disease hospitalizations--using 1999 data on 11,312 men aged 45-85 years in Malmo, Sweden.
Using a conceptual rather than a mathematical approach, this article proposed a link between multilevel regression analysis (MLRA) and social epidemiological concepts. It has been previously explained that the concept of clustering of individual health status within neighbourhoods is useful for operationalising contextual phenomena in social epidemiology. It has been shown that MLRA permits investigating neighbourhood disparities in health without considering any particular neighbourhood characteristic but only information on the neighbourhood to which each person belongs. This article illustrates how to analyse cross level (neighbourhood-individual) interactions, how to investigate associations between neighbourhood characteristics and individual health, and how to use the concept of clustering when interpreting those associations and geographical differences in health. Design and
A MLRA was performed using hypothetical data pertaining to systolic blood pressure (SBP) from 25 000 subjects living in the 39 neighbourhoods of an imaginary city. Associations between individual characteristics (age, body mass index (BMI), use of antihypertensive drug, income) or neighbourhood characteristic (neighbourhood income) and SBP were analysed.
About 8% of the individual differences in SBP were located at the neighbourhood level. SBP disparities and clustering of individual SBP within neighbourhoods increased along individual BMI. Neighbourhood low income was associated with increased SBP over and above the effect of individual characteristics, and explained 22% of the neighbourhood differences in SBP among people of normal BMI. This neighbourhood income effect was more intense in overweight people.
Measures of variance are relevant to understanding geographical and individual disparities in health, and complement the information conveyed by measures of association between neighbourhood characteristics and health.
In social epidemiology, it is easy to compute and interpret measures of variation in multilevel linear regression, but technical difficulties exist in the case of logistic regression. The aim of this study was to present measures of variation appropriate for the logistic case in a didactic rather than a mathematical way. Design and
Data were used from the health survey conducted in 2000 in the county of Scania, Sweden, that comprised 10 723 persons aged 18-80 years living in 60 areas. Conducting multilevel logistic regression different techniques were applied to investigate whether the individual propensity to consult private physicians was statistically dependent on the area of residence (that is, intraclass correlation (ICC), median odds ratio (MOR)), the 80% interval odds ratio (IOR-80), and the sorting out index).
The MOR provided more interpretable information than the ICC on the relevance of the residential area for understanding the individual propensity of consulting private physicians. The MOR showed that the unexplained heterogeneity between areas was of greater relevance than the individual variables considered in the analysis (age, sex, and education) for understanding the individual propensity of visiting private physicians. Residing in a high education area increased the probability of visiting a private physician. However, the IOR showed that the unexplained variability between areas did not allow to clearly distinguishing low from high propensity areas with the area educational level. The sorting out index was equal to 82%.
Measures of variation in logistic regression should be promoted in social epidemiological and public health research as efficient means of quantifying the importance of the context of residence for understanding disparities in health and health related behaviour.
This paper describes some of the major points of progress and challenges in health management of dairy cattle in the last 25 yr. A selection of the leading contributors in the field is acknowledged. Specific advances in the areas of transition cow management, epidemiology, udder health, applied immunology, housing design, calf health, and health-monitoring tools are described. The greatest advances in dairy health in the last 25 yr have been the shifts to disease prevention, rather than treatment, as well as from focus on individual animals to groups and herds. A fundamental advancement has been recognition of the multifactorial nature of almost all diseases of importance in dairy cattle. Epidemiology has been a critical new tool used to describe and quantify the interconnected risk factors that produce disease. Another major advance has been redefining disease more broadly, to include subclinical conditions (e.g., subclinical mastitis, ketosis, rumen acidosis, and endometritis). This expansion resulted both from improved technology to measure function at the organ level and, just as importantly, from the evolution of the health management paradigm in which any factor that limits animal or herd performance might be considered a component of disease. Links between cattle and people through consideration of environmental or ecosystem health are likely to further expand the concept of disease prevention in the future.
The decline in dairy herd fertility internationally has highlighted the limited impact of traditional veterinary approaches to herd fertility. The role of the veterinarian in fertility management on dairy farms has evolved from addressing individual clinical conditions to analyzing suboptimal herd metrics. However, this paradigm shift has only successfully occurred in some dairy industries and less so in others. Needs analyses indicate that the critical constraints to change are veterinary practice size, client motivation and data quality and availability. In addition, this review identified the inability of veterinarians to demonstrate and to market the cost-benefit of their fertility management services as important impediments to change. In many cases change is not being managed but is imposed by the growth of paraprofessionals. Some veterinarians still see their role as an animal clinician while others have evolved into leaders of the herd fertility management team. The core role of dairy veterinarians remains individual animal examinations but this must be supplemented with systematic herd fertility investigation and veterinarian-led herd fertility management. This new role encompasses leading the change from clinical calls only to a planned approach to herd fertility, demonstrating the cost-benefits of the program, scheduling fertility management consultations, assisting the farmer in setting specific, measurable, attainable, relevant and time-limited (SMART) goals, drawing up standard operating procedures (SOPs), training and auditing staff in fertility management practices, encouraging a team approach, implementing veterinary fertility management and monitoring performance. Veterinarians who fail to engage in this process of change risk being marginalized by others keen to promote their herd fertility services.
This paper reviews the causes, impact, treatment, and prevention of retained placenta (RP), metritis, and endometritis in dairy cows. The occurrence of each of these diseases largely depends on immune function in the transition period. Retained placenta affects 5-10% of calvings and greatly increases the risk of metritis and endometritis. More field studies are needed to validate criteria for treatment of metritis, but cows with at least two of RP, fever, dullness, and fetid uterine discharge appear to merit treatment with systemic antibiotics. Clinical endometritis affects 15-20% of cows at 4-6 weeks postpartum; an additional 30-35% have subclinical endometritis between 4 and 9 weeks postpartum. Under specific conditions, treatment of cows with endometritis improved pregnancy rate. Systematic use of prostaglandin F(2alpha) at 5 and 7 weeks postpartum may improve pregnancy rate. The economic benefit of efforts to identify and treat endometritis is herd-specific.
The aggregate data study design provides an alternative group level analysis to ecological studies in the estimation of individual level health risks. An aggregate model is derived by aggregating a plausible individual level relative rate model within groups, such that population-based disease rates are modelled as functions of individual level covariate data. We apply an aggregate data method to a series of fictitious examples from a review paper by Greenland and Robins which illustrated the problems that can arise when using the results of ecological studies to make inference about individual health risks. We use simulated data based on their examples to demonstrate that the aggregate data approach can address many of the sources of bias that are inherent in typical ecological analyses, even though the limited between-region covariate variation in these examples reduces the efficiency of the aggregate study. The aggregate method has the potential to estimate exposure effects of interest in the presence of non-linearity, confounding at individual and group levels, effect modification, classical measurement error in the exposure and non-differential misclassification in the confounder.
A prospective view of culling
- Eicker
Eicker, S., Fetrow, J., 2003. A prospective view of culling. In: 2003 Midwest Dairy
Herd Health Conference. University of Wisconsin-Madison, Madison.
episensr: Basic sensitivity analysis of epidemiological results
- D Haine
Haine, D., 2016. episensr: Basic sensitivity analysis of epidemiological results. R
package version 0.8.0. https://cran.r-project.org/web/packages/episensr/.
Dead cows tell no.. .'. In: The North American Veterinary Conference
- P Rapnicki
- S Stewart
- J Fetrow
- S Eicker
Rapnicki, P., Stewart, S., Fetrow, J., Eicker, S., 2003. 'Dead cows tell no...'. In: The
North American Veterinary Conference 2003, Large Animal., pp. 18-22 http://
ovidsp.ovid.com/ovidweb.
cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=caba6&AN=20063003867.
RStan: the R interface to Stan
- Stan Development Team
Stan Development Team, 2016. RStan: the R interface to Stan. http://mc-stan.org.
‘Dead cows tell no …’
- Rapnicki