Materials and Methods

The study protocol was reviewed and approved by the University of California-Davis Animal Care and Use Committee. A prospective cohort observational study was carried out on a feedlot in central California in fall 2004; informed consent from the owner of the animals was obtained prior to starting the study. The 900 dairy-breed steer and bull calves enrolled in the study arrived at the feedlot in 2 shipments of 450 calves on 2 days (21-day interval between shipments). Calves originated from a single calf ranch, which bought day-old bull calves from several local dairies. At entry into the feedlot, the calves were 3 to 4 months old. A 10-mL blood sample was collected (by BRH or ACM) from each calf at that time; after clotting, blood samples were centrifuged on-site to provide serum for analysis. Calves were individually identified, weighed, and processed according to a standard feedlot protocol, which included castrating bulls (3 calves arrived as bulls; the others had been castrated at the calf ranch) and vaccinating all calves against infectious bovine rhinotracheitis virus, BVDV, bovine respiratory syncytial virus, parainfluenza-3 virus, Mannheimia haemolytica, and Clostridia spp. Each group of 450 calves remained together in a single feedlot pen. Only study calves were kept in these 2 pens, which were adjacent to each other. Other groups of cattle had fenceline contact on either side of the study pens. A second 10-mL blood sample was obtained (by BRH or ACM) from all surviving calves at 1 time point, which was either 149 or 170 days after entry to the feedlot. At this time, calves were weighed, allowing for calculation of individual average daily gain (second weight minus initial weight divided by number of days that the calf received feed).

Serum samples were stored and transported on ice and were submitted to the California Animal Health and Food Safety Laboratory, Davis branch within 6 hours of collection. Sera were assessed for antibodies against N caninum the following day by use of a kinetic ELISA.¹³ The S:P ratio was reported for all samples. Briefly, the optical density of the sample is compared to that of the positive control sample, and a ratio is created. A ratio greater than a predefined cutoff is considered positive. Bovine viral diarrhea virus infection was detected by use of a commercially available antigen capture ELISA.^a Calves that had BVDV infection at entry to the feedlot were retested at the second visit to confirm that calves had PI with BVDV.

Calf health was monitored daily by feedlot personnel. Calves suspected of being ill were moved to an animal handling facility and treated according to a standard protocol. Calves that died underwent a postmortem examination, and cause of death was assigned. The disease status of a calf was defined as chronic if it was treated ≥ 3 times for disease on separate occasions. Individual health records were maintained, including diagnosis, treatments, and outcome; only selected data were analyzed in this study.

Statistical analysis was performed by use of commercially available software.^b,c Prevalence of infection was analyzed separately by agent and by pen. Prevalence of illness and death for calves with positive or negative test results was calculated by pen. Differences in initial weight, final weight, and average daily gain between calves with positive and negative test results were compared on a pen-level basis with a 2-sided t test. The association between infection status and morbidity events and between infection status and mortality rate were examined by use of a stratified 2 X 2 table. If the P value for the homogeneity of odds ratio test was > 0.05, a common odds ratio across pens was calculated; otherwise, separate results for each pen were presented. A Cox regression analysis was performed to determine differences in time to first morbidity event between calves with positive and negative test results (with pen as a covariate). For all comparisons, P < 0.05 was considered significant.

Results

The proportion of cattle that were seropositive for N caninum at entry to the feedlot for pens 1 and 2 was 16.9% and 16.4%, respectively. There was no significant difference in seroprevalence between the 2 pens, so results were combined to give an overall seroprevalence of 16.7% (150/900 calves; 95% CI, 14.3% to 19.3%). At the second sample collection approximately 5 months later, seroprevalence was 15.9% (133/836 calves; 95% CI, 13.5% to 18.6%; Table 1). Again, there was no significant difference in seroprevalence between pens. The distribution of S:P ratios for all samples was calculated (an S:P ratio > 0.45 is reported as positive for N caninum¹³; Figure 1).

View Full Size

Distribution of S:P ratios for serum antibody against Neospora caninum among 900 dairy-breed steer calves at entry (gray bars) to a California feedlot and approximately 5 months later (black bars).

Citation: Journal of the American Veterinary Medical Association 230, 7; 10.2460/javma.230.7.1038

• Download Figure
• Download figure as PowerPoint slide

The effect of serologic status for N caninum on calf health variables was assessed (Table 2). Calves in pen 1 were 4.5 (95% CI, 2.6 to 7.9) times as likely to be ill at arrival, 1.7 (95% CI, 1.3 to 2.2) times as likely to be treated after arrival, and 2.8 (95% CI, 1.0 to 7.9) times as likely to develop chronic disease as calves in pen 2. There was no difference in overall mortality rate between pens (odds ratio, 1.2; 95% CI, 0.7 to 2.1). After stratifying by pen, there was no significant difference between seronegative and seropositive calves with respect to detectable illness on arrival, treatment after arrival, or overall chronicity; therefore, common odds ratios were calculated for these variables. However, in pen 1, the overall mortality rate was significantly greater for seropositive calves than it was for seronegative calves, whereas in pen 2, the converse was true.

Time to first detected morbidity event was evaluated via Cox regression analysis (Figure 2). There was a significant (P < 0.001) pen effect. The hazard ratio for pen was 1.8 (95% CI, 1.5 to 2.3), indicating that at any point in time, the risk of a calf having a morbidity event was 1.8 times as great for calves in pen 1 as it was for calves in pen 2. No significant differences in time to first morbidity event were evident between seropositive and seronegative calves in either pen.

View Full Size

Cox regression survival curves for days until first treatment for 2 pens of dairy-breed steers (450 calves each) that were seropositive or seronegative for N caninum.

Citation: Journal of the American Veterinary Medical Association 230, 7; 10.2460/javma.230.7.1038

• Download Figure
• Download figure as PowerPoint slide

The effect of serologic status on weight and average daily gain was dependent on pen (Table 3). In pen 2, there was no significant difference in initial weight, final weight, or average daily gain between seropositive and seronegative calves. However, in pen 1, seronegative calves weighed 12.8 kg (28.2 lb) more at the second sample collection (P = 0.005) and had a 0.05 kg/d (0.1 lb/d) advantage in average daily gain (P = 0.02), compared with seropositive calves.

Results of the antigen capture ELISA for BVDV were positive for 3 calves at entry to the feedlot (0.3%; 95% CI, 0.07% to 1.0%): 2 were in pen 1, and 1 was in pen 2. All 3 calves did not have serum anti–N caninum antibodies. One of the BVDV-positive calves from pen 1 died (cause of death reported as liver abscesses) 2 months after entry to the feedlot. The other 2 BVDV-positive calves remained healthy throughout the trial, and both had positive BVDV test results again at the second sample collection, confirming PI with BVDV (BVDV PI prevalence, 0.2% [2/836 calves]; 95% CI, 0.03% to 0.9%). These 2 calves had average daily gains of 1.16 kg/d (2.5 lb/d) and 0.79 kg/d (1.7 lb/d).

Discussion

Results of the present study have indicated that the prevalence of N caninum infection among dairy-breed steer calves in a feedlot in California was greater than that reported from beef-breed steer calves in Texas, Alberta, and Iowa^7–9; in those regions, seroprevalence of N caninum infection at entry to the feedlot was 13%,⁷ 6.5%,⁸ and 5.9%,⁹ respectively. We were not able to determine the source of calves in our study; however, the calf ranch that provides calves to the feedlot is located in an area of California where the mean number of milking cows per herd is > 1,000. Those dairies have a drylot design in which feed is provided in bunks adjacent to the pens. Such animal density may provide greater opportunity for horizontal transmission of N caninum through contaminated feed or environment.

No relationship between serologic status for N caninum at entry to the feedlot and risk of illness on arrival, receiving treatment after arrival, or overall disease chronicity was detected in calves of the present study. This is in agreement with results of other feedlot studies.^7–9 Marked differences between pens with regard to risk of treatment and risk of chronicity were evident. Compared with pen 2, pen 1 had significantly more calves that required treatment (either at or after arrival) and had more calves that were culled because of chronic illness. Morbidity rates can vary widely among pens of feedlot cattle (from 15% to ≥ 45%), depending on exposure to viral and bacterial infectious agents, intrinsic host defenses, and environmental stressors.¹⁴ It is possible that the calves in pen 1 had either greater exposure or increased stress during the immediate period before arrival to the feedlot, which led to increased illness in this group.

In the present study, the effect of serologic status for N caninum on growth rates was dependent on pen. Although seropositive calves in pen 1 weighed less at the second sample collection and had lower average daily gain than seronegative animals, there was no difference in weight or average daily gain between seropositive and seronegative calves in pen 2. These differences between pens may be associated with their significantly different rates of morbidity. Other researchers have hypothesized that stressful conditions, such as weaning or respiratory tract illness, may result in dissemination of N caninum tachyzoites to various body systems, resulting in the development of inflammation and impairment of body function.⁷ This could help explain the finding that the pen with higher morbidity rate had reduced growth in seropositive calves, even though there was no within-pen difference in morbidity rate between seropositive and seronegative calves. This is also consistent with the finding that the risk of death for seropositive calves in pen 1 was 2.4 times as great as that for seronegative calves in pen 1.

It is interesting to compare results of our study with other investigations^7,8 that have examined the effect of infection with N caninum on calf growth rates. Compared with seronegative calves, a reduction in average daily gain was detected in seropositive calves in a study⁷ of 1,009 calves from Texas (0.05 kg/d less for seropositive calves), as well as in a Canadian study⁸ of 1,976 calves (0.04 kg/d [0.09 lb/d] less for seropositive calves), although the latter reduction was not significant. In the present study, there was a significant reduction in average daily gain of 0.05 kg/d in seropositive calves in 1 of 2 pens. Given the similarity in numerical difference in average daily gain in all 3 studies, it would appear that N caninum has a negative impact on growth in feedlot cattle. Further studies to investigate this relationship are warranted, particularly in pens with high morbidity rates. The study⁷ performed in Texas revealed that seropositive calves were associated with a loss in income because of lighter hot carcass weight and increased cost of treatment ($15.62/calf), compared with seronegative calves.

By use of an S:P value of 0.45 as a cutoff point, it was determined that 29 seronegative calves became seropositive during the interval between sample collections and that a crude incidence of seroconversion was 8.3/100 calf-years at risk. The California Animal Health and Food Safety Laboratory reports results of N caninum testing as probabilities of infection.¹⁵ An S:P ratio between 0.20 and 0.45 is consistent with a 9% probability of infection, and the result is interpreted as negative. An S:P ratio > 0.45 but ≤ 0.70 corresponds to a 68% probability of infection, and the result is interpreted as positive. An S:P ratio > 0.70 is consistent with a 100% probability of infection, and the result is also interpreted as positive. Of the 29 calves that seroconverted, 23 had values that were between 0.45 and 0.70 and 6 had S:P ratios > 0.70 at the time of the second sample collection. If the conservative value of 6 seroconversions is considered, then the crude incidence of seroconversion becomes 1.7/100 calf-years at risk. A study⁶ of dairy heifers in the United Kingdom revealed an overall incidence of seroconversion of 1.9/100 dairy heifer-years at risk, whereas a study¹⁶ in California revealed an incidence of seroconversion of 8.5/100 pregnant dairy cows/y. Results of another study¹⁷ of 2 dairies in California indicated that the estimated postnatal infection rate was < 1%/y. Thus, it appears that seroconversion may be occurring at a low level during all phases of animal growth.

The changes in S:P ratios identified in the present study may indicate false-positive reactions, true sero-conversion, or transient serologic responses. The kinetic ELISA has a stated sensitivity of 88% and specificity of 97% when a value of 0.45 is used as the cutoff for interpretation based on a seroprevalence of 40%¹³; therefore, some of the results may represent false-positive reactions. By assuming that sensitivity and specificity are constant at various levels of prevalence, it is possible to calculate the estimated true prevalence in the study sample as 16.1%, the predictive value of a positive test is 0.849, and the predictive value of a negative test is 0.977. On the basis of these calculations, 84.9% of calves that have positive test results are truly infected and 97.7% of calves that have negative test results are truly not infected; therefore, it is likely that there were several false-positive reactions as well as a few false-negative reactions in our study. If the calves truly seroconverted as a result of postnatal acquisition of infection, the source of oocysts is uncertain. Transient serologic response to N caninum in beef steers has been previously reported¹⁸; in that study, a significant impairment in feed efficiency was detected following a positive serologic result.

Conversely, 36 of 140 (25.7%) calves that were seropositive for N caninum initially were seronegative at the second sample collection. These results may represent loss of colostral antibodies, false-negative results, or transient serologic responses. Calves in our study entered the feedlot at approximately 3 to 4 months of age. Therefore, some of the initial positive test results for N caninum could have occurred because of the acquisition of antibodies through ingestion of colostrum. Decay of anti–N caninum antibody concentration follows first-order decay kinetics with a mean half-life of 19.6 ± 5.2 days.¹⁷ Colostral antibody concentration decreases to negative S:P ratios by 128 days in 100% of calves and by 33 days in 50% of calves,¹⁶ making it unlikely that most of the conversions from seropositive to sero-negative were attributable to loss of colostral antibody. Given the imperfect specificity of the N caninum kinetic ELISA, a small number of false-negative results is possible. As stated previously, transient serologic responses have been reported,¹⁸ which may also help explain some of the seroconversions among calves in our study.

Persistent infection with BVDV was found at a low prevalence (0.2%) in the calves of our study. This is consistent with findings of other feedlot studies^9,19–21; PI was detected in 0.3% of 667 calves, 0.3% of 2,000 calves, 0.2% of 5,041 calves, and 0.4% of 21,743 calves. These data appear to indicate that PI with BVDV develops in dairy-breed steer calves at a low prevalence similar to that found in beef-breed calves.

Overall, although N caninum serostatus had no significant effect on morbidity rate or chronicity of infection among calves in the present study, seropositive calves in 1 pen that had a high morbidity rate had gained less weight at approximately 5 months after entry to the feedlot than the seronegative calves in that pen. Further investigation into the mechanistic processes behind this finding is warranted.