Risk behavior has substantial consequences for health, well-being, and general behavior. The association between real-world risk behavior and risk behavior on experimental tasks is well documented, but their modeling is challenging for several reasons. First, many experimental risk tasks may end prematurely leading to censored observations. Second, certain outcome values can be more attractive than others. Third, a priori unknown groups of participants can react differently to certain risk-levels. Here, we propose the censored mixture model which models risk taking while dealing with censoring, attractiveness to certain outcomes, and unobserved individual risk preferences, next to experimental conditions.

The linear regression model y=β′x+ ε is reanalyzed. Taking the modest position that β′x is an approximation of the “best” predictor of y we derive the asymptotic distribution of b and R2, under mild assumptions.

The method of derivation yields an easy answer to the estimation of β from a data set which contains incomplete observations, where the incompleteness is random.

An isotope dilution model for partitioning phenylalanine and tyrosine uptake by the liver of the lactating dairy cow is constructed and solved in the steady state. An original ten-pool model is adopted and solved by cleaving it into two five-pool sub-models, one representing phenylalanine and the other tyrosine. If assumptions are made, model solution permits calculation of the rate of phenylalanine and tyrosine uptake from portal vein and hepatic arterial blood supply, hydroxylation, and synthesis and degradation of constitutive protein. The model requires the measurement of plasma flow rate through the liver in combination with amino acid concentrations and plateau isotopic enrichments in arterial and portal and hepatic vein plasma during a constant infusion of [1-13C]phenylalanine and [2,3,5,6-2H]tyrosine tracers. It also requires estimates of the rate of oxidation and protein export secretion. Analysis of measurement errors in experimental enrichments and infusion rates on model solutions indicated that accurate values of the intracellular and extracellular enrichments are central to minimising errors in the calculated flows. Solving the model by cleaving into two five-pool schemes rather than solving the ten-pool scheme directly is preferred as there appears to be less compounding of errors and the results consistently appear to be more biologically feasible. The model provides a means for assessing the impact of hepatic metabolism on amino acid availability to peripheral tissues such as the mammary gland.

Exposure to pregnancy complications, including preeclampsia (PE), has lifelong influences on offspring’s health. We have previously reported that experimental PE, induced in mice by administration of adenoviral sFlt1 at gestational day 8.5 combined with LPS at day 10.5, results in symmetrical growth restriction in female and asymmetrical growth restriction in male offspring. Here, we characterize the molecular phenotype of the fetal brain and liver with respect to gene transcription and DNA methylation at the end of gestation.

In fetal brain and liver, expression and DNA methylation of several key regulatory genes is altered by PE exposure, mostly independent of fetal sex. These alterations point toward a decreased gluconeogenesis in the liver and stimulated neurogenesis in the brain, potentially affecting long-term brain and liver function. The observed sex-specific growth restriction pattern is not reflected in the molecular data, showing that PE, rather than tissue growth, drives the molecular phenotype of PE-exposed offspring.

Phenylalanine (PHE) and to a lesser extent TYR are two commonly used amino acid tracers for measuring protein metabolism in a variety of species and tissues. The model examined in this paper was developed to resolve trans-organ and stable isotope dilution data collected from experiments with lactating dairy cows using these tracers. Two methods of solving the model, i.e. as two four-pool submodels, one representing PHE and the other TYR, or as an integrated eight-pool model, are investigated and the alternative solutions are contrasted. Solving the model as the two four-pool submodels rather than the integrated 8-pool model is preferred as the equations are slightly simpler and their application less susceptible to any compounding of measurement errors. The data used to illustrate the model were taken from experiments conducted to investigate the effects of high and low protein diets on the partitioning of PHE and TYR between milk protein synthesis and other metabolic fates by the mammary gland.

Background: For the rising number of people living with dementia, cost-effective community-based interventions to support psychosocial care are needed. The FindMyApps program helps people with dementia and their caregivers learn to use tablet computers and find user-friendly apps that facilitate self-management and engagement in meaningful activities. This definitive trial builds on previous feasibility pilot trials of FindMyApps and further evaluates cost-effectiveness.

Method: This is a protocol for a non-blinded randomized controlled trial (RCT) with two arms (intervention and usual care). 150 dyads (person with dementia and their carer) will be recruited. Participants must be resident in the community, with a diagnosis of Mild Cognitive Impairment or mild dementia (Mini Mental-State Examination 17-26, or Global Deterioration Scale 3-4. Dyads will be randomly assigned in equal proportions to receive either the FindMyApps intervention (experimental arm) or usual care (control arm). Primary outcomes measured at 3 months will be: patient self-management and social participation; caregiver sense of competence. Data will be collected through questionnaires filled in by the researcher (patient outcomes) or participants themselves (carer outcomes). In addition to a main effect analysis, a cost-effectiveness analysis will take place. In line with Medical Research Council (MRC) guidance for the evaluation of complex interventions, a process analysis will be undertaken, to identify factors that may influence trial outcomes. Semi-structured interviews and remotely collected data regarding use of the FindMyApps app will support the process analysis.

Result: Results of this study are expected in 2022. The study will be adequately powered to detect at least a moderate effect size of the intervention with respect to the primary outcomes.

Conclusion: This study will investigate the effectiveness and cost-effectiveness of the FindMyApps intervention. The results of the study will provide strong evidence to support or oppose scaling up implementation of the intervention. This is also an example of how the MRC framework for the evaluation of complex interventions can be implemented in practice. In a field which is often criticized for a lack of high quality evidence, randomized controlled trials should be applied more frequently designed for the robust and transparent evaluation of digital tools and technologies.

We investigate C-sets in almost zero-dimensional spaces, showing that closed $\sigma $C-sets are C-sets. As corollaries, we prove that every rim-$\sigma $-compact almost zero-dimensional space is zero-dimensional and that each cohesive almost zero-dimensional space is nowhere rational. To show that these results are sharp, we construct a rim-discrete connected set with an explosion point. We also show that every cohesive almost zero-dimensional subspace of $($Cantor set$)\!\times \mathbb R$ is nowhere dense.

In mitigating greenhouse gas (GHG) emissions and reducing the carbon footprint of dairy milk, the use of generic estimates in inventory and accounting methodology at farm level largely ignores variation of on-farm GHG emissions. The present study aimed to implement results of an extant dynamic, mechanistic Tier 3 model for enteric methane (CH4) (applied in Dutch national GHG inventory) in order to capture variation in enteric CH4 emission, and in faecal N and organic matter (OM) digestibility, ultimately required to predict manure CH4 and ammonia emission. Tier 3 model predictions were translated into calculation rules that could easily be implemented in an annual nutrient cycling assessment tool including GHG emissions, which is currently used by Dutch dairy farmers. Calculations focussed on (1) enteric CH4 emission, (2) apparent faecal OM digestibility and (3) apparent faecal N digestibility. Enteric CH4 was expressed in CH4 yield indicated with the term emission factor (EF; g CH4/kg DM) for individual dietary components and feedstuffs. Factors investigated to cover predicted variation in EF value included the level of feed intake, the type of roughage fed (proportions of grass silage and maize silage) and the quality of roughage fed. A minimum number of three classes of roughage type (i.e. 0. 40% and 80% maize silage in roughage DM) appeared necessary to obtain correspondence between interpolated EF values from EF lists and Tier 3 model predictions. A linear decline in EF value with 1% per kg increase in DM intake is adopted based on model simulations. The quality of roughage was represented by the effect of maturity of harvested grass or of the whole plant maize at cutting, based on a survey of modelling as well as experimental work. Also, predictions were assembled for apparent faecal OM digestibility which could be used in national inventory and in farm accounting. Apparent faecal N digestibility (as a major determinant of predicted urinary N excretion) was predicted, to support current Dutch national ammonia emission inventory and to correct the level of N digestibility in farm accounting. Compared to generic values or values retrieved from the Dutch feeding tables, predicted OM and N digestibility and enteric CH4 are better rooted in physiological principles and better reflect observed variation under experimental conditions. The present results apply for conditions with fairly intensive grassland management in temperate regions.

Mechanistic models (MMs) have served as causal pathway analysis and ‘decision-support’ tools within animal production systems for decades. Such models quantitatively define how a biological system works based on causal relationships and use that cumulative biological knowledge to generate predictions and recommendations (in practice) and generate/evaluate hypotheses (in research). Their limitations revolve around obtaining sufficiently accurate inputs, user training and accuracy/precision of predictions on-farm. The new wave in digitalization technologies may negate some of these challenges. New data-driven (DD) modelling methods such as machine learning (ML) and deep learning (DL) examine patterns in data to produce accurate predictions (forecasting, classification of animals, etc.). The deluge of sensor data and new self-learning modelling techniques may address some of the limitations of traditional MM approaches – access to input data (e.g. sensors) and on-farm calibration. However, most of these new methods lack transparency in the reasoning behind predictions, in contrast to MM that have historically been used to translate knowledge into wisdom. The objective of this paper is to propose means to hybridize these two seemingly divergent methodologies to advance the models we use in animal production systems and support movement towards truly knowledge-based precision agriculture. In order to identify potential niches for models in animal production of the future, a cross-species (dairy, swine and poultry) examination of the current state of the art in MM and new DD methodologies (ML, DL analytics) is undertaken. We hypothesize that there are several ways via which synergy may be achieved to advance both our predictive capabilities and system understanding, being: (1) building and utilizing data streams (e.g. intake, rumination behaviour, rumen sensors, activity sensors, environmental sensors, cameras and near IR) to apply MM in real-time and/or with new resolution and capabilities; (2) hybridization of MM and DD approaches where, for example, a ML framework is augmented by MM-generated parameters or predicted outcomes and (3) hybridization of the MM and DD approaches, where biological bounds are placed on parameters within a MM framework, and the DD system parameterizes the MM for individual animals, farms or other such clusters of data. As animal systems modellers, we should expand our toolbox to explore new DD approaches and big data to find opportunities to increase understanding of biological systems, find new patterns in data and move the field towards intelligent, knowledge-based precision agriculture systems.

Rumen sensors provide specific information to help understand rumen functioning in relation to health disorders and to assist in decision-making for farm management. This review focuses on the use of rumen sensors to measure ruminal pH and discusses variation in pH in both time and location, pH-associated disorders and data analysis methods to summarize and interpret rumen pH data. Discussion on the use of rumen sensors to measure redox potential as an indication of the fermentation processes is also included. Acids may accumulate and reduce ruminal pH if acid removal from the rumen and rumen buffering cannot keep pace with their production. The complexity of the factors involved, combined with the interactions between the rumen and the host that ultimately determine ruminal pH, results in large variation among animals in their pH response to dietary or other changes. Although ruminal pH and pH dynamics only partially explain the typical symptoms of acidosis, it remains a main indicator and may assist to optimize rumen function. Rumen pH sensors allow continuous monitoring of pH and of diurnal variation in pH in individual animals. Substantial drift of non-retrievable rumen pH sensors, and the difficulty to calibrate these sensors, limits their application. Significant within-day variation in ruminal pH is frequently observed, and large distinct differences in pH between locations in the rumen occur. The magnitude of pH differences between locations appears to be diet dependent. Universal application of fixed conversion factors to correct for absolute pH differences between locations should be avoided. Rumen sensors provide high-resolution kinetics of pH and a vast amount of data. Commonly reported pH characteristics include mean and minimum pH, but these do not properly reflect severity of pH depression. The area under the pH × time curve integrates both duration and extent of pH depression. The use of this characteristic, as well as summarizing parameters obtained from fitting equations to cumulative pH data, is recommended to identify pH variation in relation to acidosis. Some rumen sensors can also measure the redox potential. This measurement helps to understand rumen functioning, as the redox potential of rumen fluid directly reflects the microbial intracellular redox balance status and impacts fermentative activity of rumen microorganisms. Taken together, proper assessment and interpretation of data generated by rumen sensors requires consideration of their limitations under various conditions.

The Kempen system is a dairy feeding system in which diet is provided in the form of a compound feed (CF) and hay offered ad libitum. Ad libitum access to CF and hay allows cows in this system to achieve a high DM intake (DMI). Out of physiological concerns, the voluntary hay intake could be increased and the consumption pattern of CF could be manipulated to maintain proper rumen functioning and health. This study investigated the effects of an artificial hay aroma and CF formulation on feed intake pattern, rumen function and milk production in mid- to late-lactating dairy cows. Twenty Holstein–Friesian cows were assigned to four treatments in a 4 × 4 Latin square design. Diet consisted of CF and grass hay (GH), fed separately, and both offered ad libitum, although CF supply was restricted in maximum meal size and speed of supply by an electronic system. Treatments were the combination of two CF formulations – high in starch (CHS) and fibre (CHF); and two GH – untreated (UGH) and the same hay treated with an artificial aroma (TGH). Meal criteria were determined using three-population Gaussian–Gaussian–Weibull density functions. No GH × CF interaction effects on feed intake pattern characteristics were found. Total DMI and CF intake, but not GH intake, were greater (P < 0.01) in TGH treatment, and feed intake was not affected by type of CF. Total visits to feeders per day, visits to the GH feeder, visits to the CF feeder and CF eating time (all P < 0.01) were significantly greater in cows fed with TGH. Meal frequency, meal size and meal duration were unaffected by treatments. Cows fed CHF had a greater milk fat (P = 0.02), milk urea content (P < 0.01) and a greater milk fat yield (P < 0.01). Cows fed TGH had a greater milk lactose content and lactose yield (P < 0.05), and milk urea content (P < 0.01). Cows fed TGH had smaller molar proportions of acetic acid and greater molar proportions of propionic acid compared with UGH. In conclusion, treatment of GH with an artificial aroma increased CF intake and total DMI, but did not affect hay intake. Additionally, GH treatment increased the frequency of visits to both feeders, and affected rumen volatile fatty acid profile. Type of CF did not affect meal patterns, ruminal pH, nor fermentation profiles.

On-farm nutrition and management interventions to reduce enteric CH4 (eCH4) emission, the most abundant greenhouse gas from cattle, may also affect volatile solids and N excretion. The objective was to jointly quantify eCH4 emissions, digestible volatile solids (dVS) excretion and N excretion from dairy cattle, based on dietary variables and animal characteristics, and to evaluate relationships between these emissions and excreta. Univariate and Bayesian multivariate mixed-effects models fitted to 520 individual North American dairy cow records indicated dry matter (DM) intake and dietary ADF and CP to be the main predictors for production of eCH4 emissions and dVS and N excreta (g/day). Yields (g/kg DM intake) of eCH4 emissions and dVS and N excreta were best predicted by dietary ADF, dietary CP, milk yield and milk fat content. Intensities (g/kg fat- and protein-corrected milk) of eCH4, dVS and N excreta were best predicted by dietary ADF, dietary CP, days in milk and BW. A K-fold cross-validation indicated that eCH4 and urinary N variables had larger root mean square prediction error (RMSPE; % of observed mean) than dVS, fecal N and total N production (on average 24.3% and 26.5% v. 16.7%, 15.5% and 16.2%, respectively), whereas intensity variables had larger RMSPE than production and yields (29.4%, 14.7% and 14.6%, respectively). Univariate and multivariate equations performed relatively similar (18.8% v. 19.3% RMSPE). Mutual correlations indicated a trade-off for eCH4v. dVS yield. The multivariate model indicated a trade-off between eCH4 and dVS v. total N production, yield and intensity induced by dietary CP content.

This study investigated the relationships between methane (CH4) emission and fatty acids, volatile metabolites (V) and non-volatile metabolites (NV) in milk of dairy cows. Data from an experiment with 32 multiparous dairy cows and four diets were used. All diets had a roughage : concentrate ratio of 80 : 20 based on dry matter (DM). Roughage consisted of either 1000 g/kg DM grass silage (GS), 1000 g/kg DM maize silage (MS), or a mixture of both silages (667 g/kg DM GS and 333 g/kg DM MS; 333 g/kg DM GS and 677 g/kg DM MS). Methane emission was measured in climate respiration chambers and expressed as production (g/day), yield (g/kg dry matter intake; DMI) and intensity (g/kg fat- and protein-corrected milk; FPCM). Milk was sampled during the same days and analysed for fatty acids by gas chromatography, for V by gas chromatography–mass spectrometry, and for NV by nuclear magnetic resonance. Several models were obtained using a stepwise selection of (1) milk fatty acids (MFA), V or NV alone, and (2) the combination of MFA, V and NV, based on the minimum Akaike’s information criterion statistic. Dry matter intake was 16.8±1.23 kg/day, FPCM yield was 25.0±3.14 kg/day, CH4 production was 406±37.0 g/day, CH4 yield was 24.1±1.87 g/kg DMI and CH4 intensity was 16.4±1.91 g/kg FPCM. The observed CH4 emissions were compared with the CH4 emissions predicted by the obtained models, based on concordance correlation coefficient (CCC) analysis. The best models with MFA alone predicted CH4 production, yield and intensity with a CCC of 0.80, 0.71 and 0.69, respectively. The best models combining the three types of metabolites included MFA and NV for CH4 production and CH4 yield, whereas for CH4 intensity MFA, NV and V were all included. These models predicted CH4 production, yield and intensity better with a higher CCC of 0.92, 0.78 and 0.93, respectively, and with increased accuracy (C b ) and precision (r). The results indicate that MFA alone have moderate to good potential to estimate CH4 emission, and furthermore that including V (CH4 intensity only) and NV increases the CH4 emission prediction potential. This holds particularly for the prediction model for CH4 intensity.

The adaptation of dairy cows to methane (CH4)-mitigating feed additives was evaluated using the in vitro gas production (GP) technique. Nine rumen-fistulated lactating Holstein cows were grouped into three blocks and within blocks randomly assigned to one of three experimental diets: Control (CON; no feed additive), Agolin Ruminant® (AR; 0.05 g/kg dry matter (DM)) or lauric acid (LA; 30 g/kg DM). Total mixed rations composed of maize silage, grass silage and concentrate were fed in a 40 : 30 : 30 ratio on DM basis. Rumen fluid was collected from each cow at days −4, 1, 4, 8, 15 and 22 relative to the introduction of the additives in the diets. On each of these days, a 48-h GP experiment was performed in which rumen fluid from each individual donor cow was incubated with each of the three substrates that reflected the treatment diets offered to the cows. DM intake was on average 19.8, 20.1 and 16.2 kg/day with an average fat- and protein-corrected milk production of 30.7, 31.7 and 26.2 kg/day with diet CON, AR and LA, respectively. In general, feed additives in the donor cow diet had a larger effect on gas and CH4 production than the same additives in the incubation substrate. Incubation substrate affected asymptotic GP, half-time of asymptotic CH4 production, total volatile fatty acid (VFA) concentration, molar proportions of propionate and butyrate and degradation of organic matter (OMD), but did not affect CH4 production. No substrate×day interactions were observed. A significant diet×day interaction was observed for in vitro gas and CH4 production, total VFA concentration, molar proportions of VFA and OMD. From day 4 onwards, the LA diet persistently reduced gas and CH4 production, total VFA concentration, acetate molar proportion and OMD, and increased propionate molar proportion. In vitro CH4 production was reduced by the AR diet on day 8, but not on days 15 and 22. In line with these findings, the molar proportion of propionate in fermentation fluid was greater, and that of acetate smaller, for the AR diet than for the CON diet on day 8, but not on days 15 and 22. Overall, the data indicate a short-term effect of AR on CH4 production, whereas the CH4-mitigating effect of LA persisted.

Grass silage is typically fed to dairy cows in temperate regions. However, in vivo information on methane (CH4) emission from grass silage of varying quality is limited. We evaluated the effect of two rates of nitrogen (N) fertilisation of grassland (low fertilisation (LF), 65 kg of N/ha; and high fertilisation (HF), 150 kg of N/ha) and of three stages of maturity of grass at cutting: early maturity (EM; 28 days of regrowth), mid maturity (MM; 41 days of regrowth) and late maturity (LM; 62 days of regrowth) on CH4 production by lactating dairy cows. In a randomised block design, 54 lactating Holstein–Friesian dairy cows (168±11 days in milk; mean±standard error of mean) received grass silage (mainly ryegrass) and compound feed at 80 : 20 on dry matter basis. Cows were adapted to the diet for 12 days and CH4 production was measured in climate respiration chambers for 5 days. Dry matter intake (DMI; 14.9±0.56 kg/day) decreased with increasing N fertilisation and grass maturity. Production of fat- and protein-corrected milk (FPCM; 24.0±1.57 kg/day) decreased with advancing grass maturity but was not affected by N fertilisation. Apparent total-tract feed digestibility decreased with advancing grass maturity but was unaffected by N fertilisation except for an increase and decrease in N and fat digestibility with increasing N fertilisation, respectively. Total CH4 production per cow (347±13.6 g/day) decreased with increasing N fertilisation by 4% and grass maturity by 6%. The smaller CH4 production with advancing grass maturity was offset by a smaller FPCM and lower feed digestibility. As a result, with advancing grass maturity CH4 emission intensity increased per units of FPCM (15.0±1.00 g CH4/kg) by 31% and digestible organic matter intake (33.1±0.78 g CH4/kg) by 15%. In addition, emission intensity increased per units of DMI (23.5±0.43 g CH4/kg) by 7% and gross energy intake (7.0±0.14% CH4) by 9%, implying an increased loss of dietary energy with advancing grass maturity. Rate of N fertilisation had no effect on CH4 emissions per units of FPCM, DMI and gross energy intake. These results suggest that despite a lower absolute daily CH4 production with a higher N fertilisation rate, CH4 emission intensity remains unchanged. A significant reduction of CH4 emission intensity can be achieved by feeding dairy cows silage of grass harvested at an earlier stage of maturity.

The in situ degradation of the washout fraction of starch in six feed ingredients (i.e. barley, faba beans, maize, oats, peas and wheat) was studied by using a modified in situ protocol and in vitro measurements. In comparison with the washing machine method, the modified protocol comprises a milder rinsing method to reduce particulate loss during rinsing. The modified method markedly reduced the average washout fraction of starch in these products from 0.333 to 0.042 g/g. Applying the modified rinsing method, the fractional degradation rate (k d ) of starch in barley, oats and wheat decreased from on average 0.327 to 0.144 h−1 whereas for faba beans, peas and maize no differences in k d were observed compared with the traditional washing machine rinsing. For barley, maize and wheat, the difference in non-fermented starch in the residue between both rinsing methods during the first 4 h of incubation increased, which indicates secondary particle loss. The average effective degradation of starch decreased from 0.761 to 0.572 g/g when using the new rinsing method and to 0.494 g/g when applying a correction for particulate matter loss during incubation. The in vitro k d of starch in the non-washout fraction did not differ from that in the total product. The calculated ratio between the k d of starch in the washout and non-washout fraction was on average 1.59 and varied between 0.96 for oats and 2.39 for maize. The fractional rate of gas production was significantly different between the total product and the non-washout fraction. For all products, except oats, this rate of gas production was larger for the total product compared with the non-washout fraction whereas for oats the opposite was observed. The rate of increase in gas production was, especially for grains, strongly correlated with the in vitro k d of starch. The results of the present study do not support the assumption used in several feed evaluation systems that the degradation of the washout fraction of starch in the rumen is much faster than that of the non-washout fraction.