Introduction
On cow–calf operations, calves sold at weaning are producers’ main revenue source (Chenoweth and Sanderson, Reference Chenoweth and Sanderson2005). Therefore, ensuring calf survival during the preweaning period is economically essential. In western Canada, 25% of herds had 2.8% and 5.3% calf mortality from 24 hours after birth until weaning in calves born from cows and heifers, respectively (Waldner et al., Reference Waldner, Parker and Campbell2019). Calf mortality is associated with calf morbidity in herds, meaning that calves that get sick have higher odds of dying (Busato et al., Reference Busato, Steiner, Martin, Shoukri and Gaillard1997; Ganaba et al., Reference Ganaba, Bigras-Poulin, Bélanger and Couture1995; Mõtus et al., Reference Mõtus, Viltrop and Emanuelson2018; United States Department of Agriculture Animal and Plant Health Inspection Service Veterinary Services National Animal Health Monitoring System, 2021). The two most important causes of morbidity before weaning are neonatal calf diarrhea (NCD) and bovine respiratory disease (BRD) (Murray et al., Reference Murray, Fick, Pajor, Barkema, Jelinski and Windeyer2016; Pearson et al., Reference Pearson, Pajor, Caulkett, Levy, Campbell and Windeyer2019a; Waldner et al., Reference Waldner, Jelinski and McIntyre-Zimmer2013). Given this, evidence-informed health management is essential to ensure that the recommended practices are being used in herds to prevent disease and thus minimize mortality.
Direct disease control practices target disease by minimizing the contact between pathogens and hosts and enhancing antigen-specific immunity (Brandt et al., Reference Brandt, Sanderson, DeGroot, Thomson and Hollis2008; Thrusfield and Christley, Reference Thrusfield and Christley2018), for example, vaccination and biosecurity (Chenoweth and Sanderson, Reference Chenoweth and Sanderson2005; Tizard, Reference Tizard2021). Indirect disease control practices manage factors that trigger disease. An example is reducing stress by using non-abrupt weaning methods (Griebel et al., Reference Griebel, Hill and Stookey2014; Hulbert and Moisá, Reference Hulbert and Moisá2016; Moggy et al., Reference Moggy, Pajor, Thurston, Parker, Greter, Schwartzkopf-Genswein, Campbell and Windeyer2017). Furthermore, various management practices are known to impact calf morbidity and mortality by mitigating or exacerbating the risk of these outcomes. For example, introducing more than 10 bulls was associated with an increased risk of BRD outbreaks (Wennekamp et al., Reference Wennekamp, Waldner, Parker, Windeyer, Larson and Campbell2021), and theoretically, based on feedlot cattle, quarantining these animals could have decreased the risk (Santinello et al., Reference Santinello, Diana, De Marchi, Scali, Bertocchi, Lorenzi, Alborali and Penasa2022).
There is a scarcity of evidence to guide health management recommendations for beef herds to prevent calf mortality. The effectiveness of practices has been mostly studied and reviewed for dairy calves (Dubrovsky et al., Reference Dubrovsky, Van Eenennaam, Karle, Rossitto, Lehenbauer and Aly2019; Godden, Reference Godden2008; Olson et al., Reference Olson, Papasian and Ritter1980; Robison et al., Reference Robison, Stott and DeNise1988; Windeyer et al., Reference Windeyer, Leslie, Godden, Hodgins, Lissemore and LeBlanc2014) and feedlot cattle (O’Connor et al., Reference O’Connor, Hu, Totton, Scott, Winder, Wang, Wang, Glanville, Wood, White, Larson, Waldner and Sargeant2019). Differences in these production systems do not allow for direct extrapolation of results to beef cow–calf operations. Therefore, there is a knowledge gap regarding the recommended practices to use in beef cow–calf herds, and the existing information has not been previously summarized. This leads to the overall question: What is the effectiveness of management practices to prevent beef calf mortality during the preweaning stage?
The objective was to assess and summarize the evidence regarding the effectiveness of disease control strategies in preventing calf mortality in beef cow–calf herds. A secondary objective was to assess the generalizability of this evidence to cow–calf operations in western Canada.
Materials and methods
This study was informed by O’Connor and Sargeant’s articles on conducting systematic reviews in veterinary medicine (O’Connor et al., Reference O’Connor, Anderson, Goodell and Sargeant2014; O’Connor and Sargeant, Reference O’Connor and Sargeant2014; Sargeant et al., Reference Sargeant, Kelton and O’Connor2014a, Reference Sargeant, Kelton and O’Connor2014b; Sargeant and O’Connor, Reference Sargeant and O’Connor2014). It is reported according to the guidelines for preferred reporting items for systematic reviews and meta-analyses (PRISMA 2020s) (Page et al., Reference Page, McKenzie, Bossuyt, Boutron, Hoffmann, Mulrow, Shamseer, Tetzlaff, Akl, Brennan, Chou, Glanville, Grimshaw, Hróbjartsson, Lalu, Li, Loder, Mayo-Wilson, McDonald, McGuinness, Stewart, Thomas, Tricco, Welch, Whiting and Moher2021).
Protocol and registration
Before the systematic review was conducted, a protocol was developed following the PRISMA-P guidelines (Moher et al., Reference Moher, Shamseer, Clarke, Ghersi, Liberati, Petticrew, Shekelle and Stewart2015). It was published in the Digital Repository of the University of Calgary (https://prism.ucalgary.ca) and online with Systematic Reviews for Animals and Food (http://www.syreaf.org/) (Sanguinetti et al., Reference Sanguinetti, Ganshorn, Agbese and Windeyer2021). After this publication, minor amendments were made, mostly related to the risk of bias (ROB) assessment (Supplementary material 1).
Eligibility criteria
The eligibility criteria were specified for the population (P), intervention (I), comparators (C), outcome (O), and study design (S) (O’Connor et al., Reference O’Connor, Anderson, Goodell and Sargeant2014).
Population
The population of interest was preweaned beef calves. Bos taurus or Bos indicus and their hybrids were included. Studies that described postweaning beef calves, feedlot, stocker, veal, dual purpose, or dairy animals were excluded.
Interventions and comparators
The interventions of interest were biosecurity and biocontainment, vaccination, colostrum management, breeding and calving season management, and nutritional management practices. These practices could be applied to pregnant dams or preweaned beef calves. Studies were required to have a concurrent comparison group (e.g. placebo or alternate management practice).
Outcome
The outcome of interest was general mortality, which included all calf deaths regardless of the cause. Studies were included if they explicitly removed stillbirths and assessed calf mortality for at least the three first months of life.
Study designs and report characteristics
Randomized and non-randomized controlled trials (RCTs and CTs) and observational studies reporting naturally occurring diseases were included. The studies were required to statistically assess the relationship between a management practice and calf mortality. The full text had to be written in English and published in a peer-reviewed journal or thesis.
Information sources
The electronic databases used for the literature search were CAB Abstracts, MEDLINE on the Ovid platform, Web of Science, and ProQuest Dissertations. The initial searches were carried out on the same day (20 May 2021) and updated (5 April 2023) to include recent publications (Supplementary material 2). Search results were imported into the software Covidence (Veritas Health Innovation, Melbourne, Australia), and the software removed duplicates. A reference list from other reviews was checked to ensure the search strategy was accurate (Chamorro and Palomares, Reference Chamorro and Palomares2020; Theurer et al., Reference Theurer, Larson and White2015). Four additional studies were manually included (Gamsjäger et al., Reference Gamsjäger, Haines, Lévy, Pajor, Campbell and Windeyer2023; Makoschey et al., Reference Makoschey, Bielsa, Oliviero, Roy, Pillet, Dufe, Valla and Cavirani2008, Reference Makoschey, Janssen, Vrijenhoek, Korsten and Marel2001; Schreiber et al., Reference Schreiber, Matheise, Dessy, Heimann, Letesson, Coppe and Collard2000).
Search strategy
The search strategy was performed in the databases using controlled vocabulary terms and keywords related to beef cattle, calves, and a list of diseases and pathogens of interest by a librarian with experience conducting systematic searches (H.G.). No language nor time restrictions were applied during the electronic database search.
Screening and selection process
Studies were screened in two stages by two independent reviewers. Before starting each stage, the process was pre-tested to ensure both reviewers understood the screening criteria (detailed in Supplementary material 3). During the first stage, titles and abstracts were screened. Signalling questions were used to guide this process. During the second stage, full texts were screened. Reviewers could classify studies as to “include” or “exclude” from the review. Conflicts during both stages were resolved through discussion between reviewers. If necessary, a third reviewer was included (Dohoo et al., Reference Dohoo, Martin and Stryhn2009; Dubrovsky et al., Reference Dubrovsky, Van Eenennaam, Karle, Rossitto, Lehenbauer and Aly2019; Sargeant and O’Connor, Reference Sargeant and O’Connor2020).
Data collection process
Two independent reviewers extracted the data from studies included in this review using pre-tested tables in Microsoft Excel (Microsoft Corporation, Redmond, WA, USA). During this stage, studies were anonymized by using a numeric code (Table 1). Information was extracted at the study level (e.g., authors, year of publication, study design, mortality risk or rate) and the practice assessment (PA) level. Practice assessment refers to the statistical assessment between individual practice and the outcome of interest. Each PA was identified using an alphanumeric code in accordance with the numeric code given to each study (Table 2; Supplementary material 7). Significant statistical associations or effects were considered if P ≤ 0.05. Statistically significant associations (A) or no statistically significant associations (NA) were the terms used to describe the findings of PAs from observational studies. Statistically significant effects (E) or no statistically significant effects (NE) were the terms used to describe the findings of PAs from RCTs and CTs. Results from univariable analyses were preferred to those from multivariable ones if both were reported, given concerns about a lack of independence among practices. If the effects of a PA were isolated from multivariable models, other variables included in the model were noted. Results were preferably extracted from tables. However, given concerns regarding the precision and validity of these estimates, the focus was on the directionality of results (i.e., protective or harmful) rather than the effect estimate. Conflict among reviewers was resolved in the same way as described previously.
Characteristics of studies included in a systematic review on the effect of management practices on preweaned calf mortality in beef herds
Summary of findings table and ROB assessment for management practices with significant associations or effects detected within a systematic review on the effect of management practices on preweaned calf mortality in beef herds
Risk of bias
The ROB assessment was done at the PA level. Practice assessments from RCTs and CTs were evaluated using the Rob2 tool (Sterne et al., Reference Sterne, Savović, Page, Elbers, Blencowe, Boutron, Cates, Cheng, Corbett, Eldridge, Emberson, Hernán, Hopewell, Hróbjartsson, Junqueira, Jüni, Kirkham, Lasserson, Li, McAleenan, Reeves, Shepperd, Shrier, Stewart, Tilling, White, Whiting and Higgins2019), as described in the Cochrane Review Handbook for Systematic Reviews of Interventions (Higgins et al., Reference Higgins, Thomas, Chandler, Cumpston, Li, Page and Welch2024). The ROBINS I tool was used to evaluate practice assessments from observational studies (Sterne et al., Reference Sterne, Hernán, Reeves, Savović, Berkman, Viswanathan, Henry, Altman, Ansari, Boutron, Carpenter, Chan, Churchill, Deeks, Hróbjartsson, Kirkham, Jüni, Loke, Pigott, Ramsay, Regidor, Rothstein, Sandhu, Santaguida, Schünemann, Shea, Shrier, Tugwell, Turner, Valentine, Waddington, Waters, Wells, Whiting and Higgins2016). A few signalling questions were modified to be applicable in veterinary medicine (Sargeant and O’Connor, Reference Sargeant and O’Connor2014), and the details are shown in Supplementary materials 4 and 5.
Data synthesis
The evidence regarding general calf mortality was summarized using a narrative structure, while evidence regarding NCD and BRD-specific morbidity and mortality (Part 2) are reported elsewhere (Sanguinetti et al., Reference Sanguinetti, Strong, Agbese, Adams, Campbell, Checkley, Ganshorn and Windeyer2025). Firstly, a summary of findings table was compiled for all PAs. If the body of evidence for a specific practice included three or more PAs from at least three different studies, a GRADE approach was used to assess the certainty of the findings (Schünemann et al., Reference Schünemann, Brożek, Guyatt and Oxman2013) (Supplementary material 6). This assessment evaluated consistency among the direction of findings across studies (i.e., beneficial or harmful), comparability of practices and comparison groups, as well as if the geography and production conditions in which the studies were conducted were comparable to those of cow–calf operations in western Canada.
Results
The search strategy identified 4942 relevant studies of which 1480 duplicates were deleted. This left 3462 studies that underwent title and abstract screening, and 3247 were excluded during this stage. The remaining 215 studies were eligible for full-text screening, and 198 were subsequently excluded (Figure 1). In total, 25 studies were retained for Parts 1 and 2 of this systematic review.
PRISMA flowchart of a systematic review on the effect of management practices on preweaned calf mortality and morbidity in beef herds.
Eleven studies were deemed relevant for the general mortality review (Part 1). These included one RCT and ten observational cross-sectional studies (Table 1). Eight took place in North America (USA and Canada), one in Europe (Estonia), one in Asia (Japan), and one in South America (Brazil). Seven out of eleven studies reported a specific case definition for mortality. The number of practices assessed in relation to mortality by each study ranged from 1 to 19, and the outcome of each practice assessment is detailed in Table 2 and Supplementary material 7.
Practices with statistically significant effects or associations reported
Colostrum practices
Three out of four PAs found that criteria used to intervene with a colostrum management strategy affected calf mortality (A: 4a, 4b, 4c; NA: 1e (Table 2)). Checking the fullness of the udder impacted the calf mortality risk from 1 to 7 days of age; herds that used this criterion had 0.7% lower mortality than those that did not (P = 0.01) (4c). Also, intervening with colostrum consumption for calves that required assistance at birth had a similar impact; herds that used this criterion had 0.8% less mortality than those that did not (P = 0.02) (4a). Herds that intervened with colostrum in the case that colostrum was abnormal had 1.9% higher mortality than those who did not (P = 0.001) (4b). Regardless of the criteria used to intervene, the findings concerning the timing to implement a colostrum management strategy (e.g., 4 hours, 12 hours after birth) (6a and 9a), the source of colostrum used to intervene (e.g., dairy) (1 f, 6b, and 6c), and the methods used to administer colostrum (e.g., using an esophagus tube) (1 g, 1 h, and 9a) did not show an impact on the odds nor rate of calf mortality (Supplementary material 7). In contrast, PA 11a reported that requiring intervention with colostrum consumption was associated with higher odds of mortality in calves (P < 0.0001, Table 2). The certainty of the evidence for colostrum practices could not be assessed, given the differences in the practices evaluated.
Timing of the calving season
Three out of four PAs reported a significant association between the timing of the year when calving took place and mortality (A: 2c, 4d, 5a; NA: 6e (Table 2)). These studies were all conducted in North America. Early calving herds had a 1.4 times higher incidence of mortality than those calving later (PA 2c). The mortality was 0.7% lower when the calving started in April (later) compared to January or February (earlier) (P = 0.02) (4d). Herds who calved earlier (January/February) had higher preweaning mortality (P = 0.02) for calves born to cows (1.9%) compared to later (March to May) calving herds (1.8%); however, this association was not detected for heifers (5a). The GRADE approach was used to assess the certainty of these findings, and it was determined that certainty was low (Table 3). Furthermore, two other studies assessed the proportion of calvings during each season as it related to calf-level outcomes and found similar results. For example, calf mortality was significantly lower in herds with a higher proportion of calvings in summer (June, July, and August) compared to those in autumn (September, October, and November), spring (March, April, and May), and winter (December, January, and February) (P < 0.001) (1a, 1b, 1c, and 1d). Also, calves born in winter and autumn had significantly higher odds of mortality than calves born in summer. There was no significant difference in mortality between summer and spring born calves (8a). However, comparisons between these two studies should be made cautiously, given that herd-level mortality was reported in one study (1a, 1b, 1c, and 1d) and calf-level mortality in the other (8a). The PAs are reported in Table 2.
Assessment of the certainty of the findings of management practices with significant effects or associations using the GRADE approach within a systematic review on the effect of management practices on preweaned calf mortality in beef herds
Length of the calving season
Two of four PAs that evaluated the length of the calving season found that it impacted calf mortality (A: 4e, 3b; NA: 1aa, 2b (Table 2)). The longer the calving season, the higher the mortality risk. The mortality from 7 days to weaning increased by 1.4% for every additional week of the calving season (P = 0.007) (4e). Similarly, another study reported a significant P-value (P = 0.002) (PA 3b), but no specific details were provided about its magnitude. The certainty of this body of evidence was determined to be low (Table 3).
Nutritional management and mineral supplementation in calves
One out of four PAs assessing the use of minerals (e.g., selenium) and/or vitamins (e.g., vitamin E) found an association with calf mortality (A: 6f; NA: 1v, 1w, 4f (Table 2)). One study reported that herds that did not use vitamin E and selenium at birth had 10.3 higher odds of mortality than those who did use vitamin E and selenium (p = 0.003) (PA 6f). Feeding minerals (1v) or selenium supplements to calves (PA 1w) or giving mineral/vitamin injections (4f) had no associations with mortality.
Practices with no statistically significant effects or associations reported
Breeding and calving management
Breeding heifers before cows was not associated with the odds of calf mortality (2a; Supplementary material 7).
Nutritional management and mineral supplementation
Using either “feeding houses” or feeding concentrates to calves did not influence the calf mortality rate (1t and 1u). Pre-calving practices, including feeding silage or giving mineral injections to dams repeatedly, were also not associated with the calf mortality risk (2d and 10a). The type of pastures, described by the authors as cultural, seminatural, natural, cultural combined, seminatural, and natural pastures, used for the cow–calf pair did not impact the calf mortality rate (1z). The PAs are shown in Supplementary material 7.
Biosecurity
Neither biocontainment nor biosecurity practices affected the calf mortality risk or rate (1i, 1j, 1r, 1s, 1x, 6d, and 6h). Biocontainment practices included disinfection of the navel cord of the newborn calf, using pastures not used for grazing in the previous year for cows and calves, grazing cows and calves separately from other animal groups, separating sick animals, removing calves from the calving facility to nursery pasture within 48 h of birth, and length of time separating calf and dam from other animals after calving. The only biosecurity practice assessed was the purchasing of foster calves, and this was not associated with mortality. Details are shown in Supplementary material 7.
Dam vaccination
The use of pre-calving vaccines against NCD pathogens were not associated with calf mortality (6g, 7b, 7c) (Supplementary material 7). The certainty of this body of evidence was determined to be low (Supplementary material 8).
Practices removed from the review
Nutritional management and mineral supplementation
Feeding cows silage (2f) was excluded from the final narrative review, because it was not specified whether this was done pre-calving or post-calving.
Vaccination
A non-inferiority trial comparing two different intranasal vaccines in calves was excluded from this review (Masset et al., Reference Masset, Meurens, Marie, Lesage, Lehébel, Brisseau and Assié2020). Both vaccines targeted Parainfluenza Virus Type 3 and Bovine Respiratory Syncytial Virus. Their differences included the strains used, tissue culture infectious doses, diluent, administration modalities, and dose of the vaccine. Vaccine A was not significantly inferior to Vaccine B (P = 0.11) in preventing calf mortality. Nonetheless, non-inferiority does not provide direct evidence about vaccination as an effective strategy to prevent calf mortality. Two other PAs were removed because of a lack of details regarding the production group that was vaccinated, disease targeted, type of vaccines used, or time of vaccination (1y and 7a).
Risk of bias assessment
Of the 42 PAs from observational cross-sectional studies and one PA from an RCT included, 22 had a ‘high’, 11 showed ‘some concerns’, and 10 had a ‘low’ ROB (Supplementary materials 9 and 10).
Twenty-three PAs were subject to selective reporting. For example, univariable analysis was not shown or only interventions with significant effects included in their final multivariable models were reported (4b and 6e). Twenty PAs did not select participants using systematic methods (e.g., used a convenience sample; 2b and 2c). Nineteen PAs did not sufficiently specify the intervention evaluated (e.g., no definition of the criteria used to define abnormal colostrum; 4b).
Discussion
The evidence compiled for the criteria used to intervene with a colostrum management strategy, timing and length of the calving season, and vitamin and mineral supplementation in calves showed statistically significant associations with calf mortality.
Determining whether a calf needs colostrum intervention depending on whether they were assisted at calving or had not nursed from their dam has been shown to reduce calf mortality at the herd level (Murray et al., Reference Murray, Fick, Pajor, Barkema, Jelinski and Windeyer2016). This aligns with the findings of an expert consensus study (Sanguinetti et al., Reference Sanguinetti, Adams, Campbell, Checkley and Windeyer2024). Assisted calves are more likely to not consume colostrum by themselves within 4 hours after birth compared to unassisted ones (Homerosky et al., Reference Homerosky, Timsit, Pajor, Kastelic and Windeyer2017). These calves also have less vigour than unassisted calves (Pearson et al., Reference Pearson, Homerosky, Caulkett, Campbell, Levy, Pajor and Windeyer2019b) and rely on colostrum intervention practices to increase their odds of survival (Besser and Gay, Reference Besser and Gay1994). Intake of colostrum that contains maternal antibodies in a timely manner is essential because calves are born with a naïve adaptive immune system and lack their own circulating antibodies (Chase et al., Reference Chase, Hurley and Reber2008; Godden, Reference Godden2008; Larson and Tyler, Reference Larson and Tyler2005; Windeyer and Gamsjäger, Reference Windeyer and Gamsjäger2019). The maternal antibodies protect the newborn as their immune system matures, thus impacting health and survival (Waldner and Rosengren, Reference Waldner and Rosengren2009). Regardless, at the calf level, calves that received any colostrum practice including different methods and sources of colostrum, had higher odds of dying than those that did not (Gamsjäger et al., Reference Gamsjäger, Haines, Lévy, Pajor, Campbell and Windeyer2023). Therefore, according to the findings of this systematic review, the decision of whether a calf needs colostrum intervention based on not nursing by themselves or being assisted at calving is the only colostrum management practice that has been shown to have an effect on calf mortality at the herd-level. However, at the individual level, these calves still have a higher risk of dying compared to those that do not require a colostrum intervention practice to be used. It is important to differentiate between individual- and herd-level practices. While some practices that have an important effect on the odds of mortality in the individual, if the practice, such as colostrum invention, is relatively rare, it may have minimal impact on the herd-level mortality risk.
Winter calving was identified as a potential risk factor for increased calf mortality (Clement et al., Reference Clement, King, Wittum, Biwer, Fleck, Salman and Odde1993; Misaka et al., Reference Misaka, Uematsu, Hashimoto, Kitahara, Osawa and Sasaki2022; Mõtus et al., Reference Mõtus, Niine, Viltrop and Emanuelson2020; Murray et al., Reference Murray, Fick, Pajor, Barkema, Jelinski and Windeyer2016; Pearson et al., Reference Pearson, Pajor, Caulkett, Levy, Campbell and Windeyer2019a). This was also described in another review (Uetake, Reference Uetake2013). Possible explanations involve the exposure of calves to cold temperatures and wind, which lower their metabolic rate and increase the time it takes to nurse from their dam (Uetake, Reference Uetake2013). At the intestinal level, this elapsed time affects the efficiency of absorbing colostrum immunoglobulins (Colazo and Kastelic, Reference Colazo and Kastelic2012; McGee and Earley, Reference McGee and Earley2019; Weaver et al., Reference Weaver, Tyler, Van Metre, Hostetler and Barrington2000). Furthermore, cold stress reduces this process even more (Olson et al., Reference Olson, Papasian and Ritter1980). Therefore, calves are more likely to have inadequate or failed transfer of passive immunity, thus increasing their risk of morbidity and mortality (Gamsjäger et al., Reference Gamsjäger, Haines, Lévy, Pajor, Campbell and Windeyer2023; Waldner and Rosengren, Reference Waldner and Rosengren2009; Windeyer et al., Reference Windeyer, Leslie, Godden, Hodgins, Lissemore and LeBlanc2014). In winter calving herds, management practices used to protect newborn calves from climatic conditions may also be associated with an increased mortality risk. For example, calving in barns involves managing animals more intensively with a higher stocking density compared to animals calving on pasture (Ganaba et al., Reference Ganaba, Bigras-Poulin, Bélanger and Couture1995; Pearson et al., Reference Pearson, Pajor, Caulkett, Levy, Campbell and Windeyer2019a; Radostits, Reference Radostits1991). A higher stocking density favours a high pathogen load in the barn and may increase the risk of transmission of pathogens between calves (Assié et al., Reference Assié, Bareille, Beaudeau and Seegers2009; Doeschl-Wilson et al., Reference Doeschl-Wilson, Knap, Opriessnig and More2021). Consequently, disease incidence is affected, given its relationship with the transmission rates (Dohoo et al., Reference Dohoo, Martin and Stryhn2009; Ogut et al., Reference Ogut, LaPatra and Reno2005). However, for some herds, for example, purebred or seedstock, calving later is not a viable option because calves need to be born as early as possible in the year to be competitive in animal shows and sales.
The length of the calving season was identified as a risk factor for calf mortality (Dutil et al., Reference Dutil, Fecteau, Bouchard, Dutremblay and Paré1999; Murray et al., Reference Murray, Fick, Pajor, Barkema, Jelinski and Windeyer2016). This may be explained by the calf acting as a pathogen amplifier during the calving season (Larson and Tyler, Reference Larson and Tyler2005). Within this review, studies that showed statistically significant associations had very different lengths of calving seasons (mean = 79 days; Murray et al., Reference Murray, Fick, Pajor, Barkema, Jelinski and Windeyer2016) versus over four months long (Dutil et al., Reference Dutil, Fecteau, Bouchard, Dutremblay and Paré1999). The findings of the first study align well with recommended management practices for herds, which state that the length should be from 60 to 80 days (Chenoweth and Sanderson, Reference Chenoweth and Sanderson2005; Colazo and Kastelic, Reference Colazo and Kastelic2012; WCCCS, 2017). Hypothetically, calves born at the beginning of the season are exposed to lower doses of environmental pathogens and are often asymptomatic if infected (Larson and Tyler, Reference Larson and Tyler2005). However, as the season progresses, the dose of environmental pathogens will increase, and consequently, later-born calves often develop clinical signs of disease (Larson and Tyler, Reference Larson and Tyler2005). Therefore, herds may reduce the risk of calf mortality by limiting the duration of the calving season. Alternatively, a short calving season may be a surrogate indicator of an unmeasured collection of good management practices that reduce calfhood mortality (i.e., herds with good reproductive management may also have good health management).
Injecting vitamin E and selenium (Se) at birth in calves also reduced calf mortality (Waldner and Rosengren, Reference Waldner and Rosengren2009). Similarly, the impact of injectable supplementation with selenium and vitamin E at birth has been reported to reduce the odds of treatment of NCD in dairy calves (Leslie et al., Reference Leslie, Nelson, Godden, Duffield, Devries and Renaud2019). In contrast, an RCT assessing repeated mineral supplementation in pregnant dams included in this review did not detect a significant effect on calf mortality compared to a control group (Stokes et al., Reference Stokes, Ireland and Shike2019) nor did another controlled trial done in western Canada assessing NCD in calves (Cohen et al., Reference Cohen, King, Guenther and Janzen1991). Within these dam studies, plasma concentrations of copper, manganese, Se, and zinc in calves at birth were not different between groups (Stokes et al., Reference Stokes, Ireland and Shike2019), nor was Se in the second study (Cohen et al., Reference Cohen, King, Guenther and Janzen1991). However, different minerals were assessed in each of these PAs, so comparisons should be made cautiously. There are several possible explanations of why statistically significant associations were found when calves were supplemented but not dams. Several steps or factors may be involved for calves to benefit from the dam supplementation. These include the severity of the initial deficiency in the supplemented dams prior to supplementation and the type of the deficiency (Cohen et al., Reference Cohen, King, Guenther and Janzen1991), the age of dams (de Weyer Lm et al., Reference de Weyer Lm, Hendrick and Waldner2010; Waldner et al., Reference Waldner, McLeod, Parker and Campbell2023), the timing of supplementation during gestation, characteristics of the products used (e.g., chelated or organic) (Ahola et al., Reference Ahola, Baker, Burns, Mortimer, Enns, Whittier, Geary and Engle2004; Chenoweth and Sanderson, Reference Chenoweth and Sanderson2005; Marques et al., Reference Marques, Cooke, Rodrigues, Cappellozza, Mills, Larson, Moriel and Bohnert2016), and the doses used to supplement (Awadeh et al., Reference Awadeh, Kincaid and Johnson1998). Furthermore, specific differences in placental or colostral absorption exist (Awadeh et al., Reference Awadeh, Kincaid and Johnson1998). For example, the absorption of Se starts in utero and is stored in the fetal liver (Gooneratne and Christensen, Reference Gooneratne and Christensen1989), while vitamin E is exclusively obtained through colostrum after birth (Quigley and Drewry, Reference Quigley and Drewry1998). In short, injecting calves is a more direct method of supplementation, avoiding these intermediate steps, and having a greater impact on calf mortality. Nevertheless, the evidence to support this practice in pregnant dams or calves within this systematic review is extremely scarce. Only one study reporting a statistically significant association is insufficient to support or discourage using this practice (Lash et al., Reference Lash, VanderWeele, Haneuse and Rothman2021). Similarly, another review identified that the impact of trace mineral supplementation in dams and its impact on calfhood health needs more research (Van Emon et al., Reference Van Emon, Sanford and McCoski2020). In western Canada, this is especially important given Se deficiency was frequently detected in the liver of beef calves that died after 3 days of age and that vitamin E deficiency was common in stillbirths (Waldner and Blakley, Reference Waldner and Blakley2014), although this latter finding should be interpreted with caution as many of these calves probably did not consume colostrum (Quigley and Drewry, Reference Quigley and Drewry1998). Overall, it is important to garner more information to better understand whether vitamin and mineral supplementation programs meet the nutritional requirements of the cattle within a given herd and are effective in optimizing the production of calves.
Within the body of evidence discussed earlier, there is consistency in the directionality of findings for some practices assessed and not others (Clement et al., Reference Clement, King, Wittum, Biwer, Fleck, Salman and Odde1993; Dutil et al., Reference Dutil, Fecteau, Bouchard, Dutremblay and Paré1999; Mõtus et al., Reference Mõtus, Niine, Viltrop and Emanuelson2020; Murray et al., Reference Murray, Fick, Pajor, Barkema, Jelinski and Windeyer2016; Pearson et al., Reference Pearson, Pajor, Caulkett, Levy, Campbell and Windeyer2019a; Waldner, Reference Waldner2008; Waldner and Rosengren, Reference Waldner and Rosengren2009). Consistency among study results supports that an actual causal relationship may exist under field conditions (Dohoo et al., Reference Dohoo, Martin and Stryhn2009). However, under certain circumstances, a cause–effect relationship may exist even when the change in the practice may not always be associated with a specific change in the direction of the outcome (Lash et al., Reference Lash, VanderWeele, Haneuse and Rothman2021). For example, a practice’s effect may vary with mortality risk (Dohoo et al., Reference Dohoo, Martin and Stryhn2009). This could explain variation in the directionality of results attained between studies, as was observed among PAs examining the length of the calving season. For this practice, two of four PAs found a statistically significant association between the length of the calving season and mortality, and two did not. In studies that reported a statistically significant association (Dutil et al., Reference Dutil, Fecteau, Bouchard, Dutremblay and Paré1999), mortality risk was > 5%. In contrast, when a statistically significant association was not observed, the mean mortality risk was estimated at 1.5% (Clement et al., Reference Clement, King, Wittum, Biwer, Fleck, Salman and Odde1993). For this hypothesis to be confirmed, several well-executed RCTs are needed to do a dose–response meta-analysis (Berlin et al., Reference Berlin, Longnecker and Greenland1993).
Most of the practices assessed in this review were not reported to have a statistically significant impact on calf mortality (Clement et al., Reference Clement, King, Wittum, Biwer, Fleck, Salman and Odde1993; Mõtus et al., Reference Mõtus, Niine, Viltrop and Emanuelson2020; Pires et al., Reference Pires, Freitas, Silva, Lima, Cyrillo, Stafuzza, Lima and Paz2021; Stokes et al., Reference Stokes, Ireland and Shike2019; Waldner, Reference Waldner2008; Waldner and Rosengren, Reference Waldner and Rosengren2009). The relationship between practices and calf mortality is not often direct, and mortality may occur only after several intermediary events, including morbidity and treatment of disease, which have a more direct relationship with management practice (Digitale et al., Reference Digitale, Martin and Glymour2022; Ganaba et al., Reference Ganaba, Bigras-Poulin, Bélanger and Couture1995). Similarly, unmeasured confounding variables may bias the reported association.
A limited number of calves dying may also limit the reliability of the findings by minimizing the sample size (Button et al., Reference Button, Ioannidis, Mokrysz, Nosek, Flint, Robinson and Munafò2013; United States Department of Agriculture Animal and Plan Health Inspection Service Veterinary Services National Animal Health Monitoring System, 2021). Within this systematic review, the mortality risk of studies varied from 1.5 to 13.5%, and none of them reported considering it for the sample size calculation. Given this, in future studies, when doing the sample size calculations, the mean mortality risk should be considered (Dohoo et al., Reference Dohoo, Martin and Stryhn2009; Wang and Cheng, Reference Wang and Cheng2020). This would help ensure that observational studies may provide more reliable results (Wang and Cheng, Reference Wang and Cheng2020). Therefore, this leads to the question of whether using these practices does not affect mortality or if a type II error is present (Akobeng, Reference Akobeng2016; Dohoo et al., Reference Dohoo, Martin and Stryhn2009; Lash et al., Reference Lash, VanderWeele, Haneuse and Rothman2021).
The statistical analysis methods could have influenced the low number of practices that showed statistically significant associations. Eight out of nine cross-sectional studies used multivariable methods to analyze the data (Clement et al., Reference Clement, King, Wittum, Biwer, Fleck, Salman and Odde1993; Dutil et al., Reference Dutil, Fecteau, Bouchard, Dutremblay and Paré1999; Misaka et al., Reference Misaka, Uematsu, Hashimoto, Kitahara, Osawa and Sasaki2022; Mõtus et al., Reference Mõtus, Niine, Viltrop and Emanuelson2020; Murray et al., Reference Murray, Fick, Pajor, Barkema, Jelinski and Windeyer2016; Pires et al., Reference Pires, Freitas, Silva, Lima, Cyrillo, Stafuzza, Lima and Paz2021; Waldner, Reference Waldner2008; Waldner and Rosengren, Reference Waldner and Rosengren2009). Within these studies, variables selected to be retained in the models relied on p-values (Lash et al., Reference Lash, VanderWeele, Haneuse and Rothman2021). Yet, the biological plausibility of associations was not assessed using directed acyclic graphs (DAGs). Only one study reported the temporal criteria used to determine whether a practice was considered to potentially impact mortality (Waldner, Reference Waldner2008). One of the assumptions of multivariable models is the independence of variables (Concato et al., Reference Concato, Feinstein and Holford1993). Assuming practices used within herds are independent is questionable and most likely unreasonable, regardless of statistical attempts to detect collinearity among variables (Fox and Monette, Reference Fox and Monette1992). For example, winter calving usually takes place in more confined areas with more intensive management, such as barns, to protect newborns from hostile temperatures and increase their odds of survival (Pearson et al., Reference Pearson, Pajor, Caulkett, Levy, Campbell and Windeyer2019a). None of the studies reported univariable analysis, so the unconditional associations between individual practices and mortality were unavailable. Therefore, although odds ratios and confidence intervals were extracted from the studies, the focus was on the directionality of the findings (e.g., beneficial or harmful to mortality risk). A meta-analysis could not be done because there was not enough reliable evidence to calculate effect estimates of any given practice.
The reliability of the findings within this systematic review is low, given that the largest bodies of evidence (i.e., timing and length of the calving season and dam vaccination against NCD pathogens) had low certainty of findings (Schünemann et al., Reference Schünemann, Brożek, Guyatt and Oxman2013). The GRADE assessment incorporates the ROB in individual studies, directionality and imprecision of results, comparability between studies, and how comparable were production conditions in the studies relative to those in cow–calf operations in western Canada. Overall, individual PAs had a high ROB, and the certainty of the bodies of evidence was downgraded because many used a cross-sectional study design. Cross-sectional studies are weak sources of evidence to infer causality given that outcomes and exposures are measured at the same time, and a temporal relationship between them cannot be demonstrated (Carlson and Morrison, Reference Carlson and Morrison2009; Dohoo et al., Reference Dohoo, Martin and Stryhn2009; Sargeant et al., Reference Sargeant, Kelton and O’Connor2014a). According to the levels of evidence approach, bodies of evidence from cross-sectional studies are considered less reliable than those from RCTs (Canadian Task Force on the Periodic Health Examination, 1979; Sargeant et al., Reference Sargeant, Brennan and O’Connor2022). Because of this, under ideal circumstances, systematic reviews should include well-executed RCTs (Burns et al., Reference Burns, Rohrich and Chung2011). However, assessing some practices using RCTs may be challenging, and well-executed cohort studies can also be a good source of evidence.
This review only included studies that explicitly removed stillbirths. Fifty percent of calf mortality occurs during the first 24 hours after birth (Pearson et al., Reference Pearson, Pajor, Caulkett, Levy, Campbell and Windeyer2019a). Calves assisted at birth have an increased risk of dying during this period (Bond and Weinland, Reference Bond and Weinland1978; Ganaba et al., Reference Ganaba, Bigras-Poulin, Bélanger and Couture1995; Wittum et al., Reference Wittum, Salman, King, Mortimer, Odde and Morris1994). Stillbirths were removed to ensure that the substantial effect of assisted calving did not statistically overshadow practices with smaller effects. However, not all calves born with assistance at calving die during the first 24 hours, and studies evaluating their survival during the preweaning stage were inevitably lost because of this exclusion criteria.
Conclusions
This review filled the knowledge gap concerning the evidence about disease control management practices to prevent calf mortality in preweaned beef calves. The timing and length of the calving season, criteria used to intervene with a colostrum management practice, and use of supplementation with vitamin E and selenium in calves were reported to have statistically significant protective associations with calf mortality. Conversely, most of the studies included were observational cross-sectional studies, and the certainty of the findings was low. Overall, the findings of this review reinforce the need to design well-executed RCTs and cohort studies to estimate the effectiveness of practices, which should be combined with those of other systematic reviews to guide evidence-informed management.
Open access
