Prevalence, distribution, and risk factors of bovine tuberculosis in dairy cattle in central China

Bovine tuberculosis (bTB) is one of the priority epidemic diseases in dairy cattle in China. Continuous surveillance and evaluation of the control programs will help on improving the efficiency of bTB control policy. We designed this study to investigate both animal and herd level prevalence of bTB, as well as to determine the associated factors in dairy farms in Henan and Hubei provinces. A cross-sectional study was conducted from May 2019 to September 2020 in central China (Henan and Hubei provinces). We sampled 40 herds in Henan and six herds in Hubei via stratified systematic sampling and administrated a questionnaire consisting of 35 factors. A total of 4900 whole blood samples were collected from 46 farms, including 545 calves < six months old and 4355 cows ≥ six months old. This study demonstrated a high animal-(18.65%, 95% CI: 17.6-19.8) and herd-level (93.48%, 95%CI: 82.1-98.6) prevalence of bTB in dairy farms in central China. The Least Absolute Shrinkage and Selection Operator (LASSO) and negative binomial regression models showed that herd positivity was associated with the practice of introducing new animals (RR = 1.7, 95%CI: 1.0-3.0, p = 0.042), and changing the disinfectant water in the wheel bath at the farm entrance every three days or less (RR = 0.4, 95%CI: 0.2-0.8, p = 0.005) which reduced the odds of herd positivity. In addition, the result illustrated that testing cows with a higher age group (≥ 60 months old) (OR=1.57, 95%CI: 1.14-2.17, p = 0.006) and within the early stage of lactation (DIM=60-120 days, OR=1.85, 95%CI: 1.19-2.88, p = 0.006) and the later stage of lactation (DIM≥301 days, OR=2.14, 95%CI: 1.30-3.52, p = 0.003) could maximize the odds of detecting seropositive animals. Our results have plenty of benefit to improve bTB surveillance strategies in China and elsewhere in the world. The LASSO and the negative binomial regression models were recommended when dealing with high herd-level prevalence and high dimensional data in questionnaire-based risk studies.