Herd-level risk factors for lameness in freestall farms in the northeastern United States and California

The objective was to investigate the association between herd-level management and facility design factors and the prevalence of lameness in high-producing dairy cows in freestall herds in the northeastern United States (NE; Vermont, New York, Pennsylvania) and California (CA). Housing and management measures such as pen space, stall design, bedding type, and milking routine were collected for the high-producing pen in 40 farms in NE and 39 farms in CA. All cows in the pen were gait scored using a 1-to-5 scale and classified as clinically lame (score ≥3) or severely lame (score ≥4). Measures associated with the (logit-transformed) proportion of clinically or severely lame cows at the univariable level were submitted to multivariable general linear models. In NE, lameness increased on farms that used sawdust bedding [odds ratio (OR)=1.71; 95% confidence interval (CI)=1.06–2.76] and decreased with herd size (OR=0.94; CI=0.90–0.97, for a 100-cow increase), use of deep bedding (OR=0.48; CI=0.29–0.79), and access to pasture (OR=0.52; CI=0.32–0.85). The multivariable model included herd size, access to pasture, and provision of deep bedding, and explained 50% of the variation in clinical lameness. Severe lameness increased with the percentage of stalls with fecal contamination (OR=1.15; CI=1.06–1.25, for a 10% increase) and with use of sawdust bedding (OR=2.13; CI=1.31–3.47), and decreased with use of deep bedding (OR=0.31; CI=0.19–0.50), sand bedding (OR=0.32; CI=0.19–0.53), herd size (OR=0.93; CI=−0.89–0.97, for a 100-cow increase), and rearing replacement heifers on site (OR=0.57; CI=0.32–0.99). The multivariable model included deep bedding and herd size, and explained 59% of the variation of severe lameness. In CA, clinical lameness increased with the percentage of stalls containing fecal contamination (OR=1.15; CI=1.05–1.26, for a 10% increase), and decreased with herd size (OR=0.96; CI=0.94–0.99, for a 100-cow increase), presence of rubber in the alley to the milking parlor (OR=0.46; CI=0.28–0.76), distance of the neck rail from the rear curb (OR=0.97; CI=0.95–0.99, for a 1-cm increase), water space per cow (OR=0.92; CI=0.85–0.99, for a 1-cm increase), and increased frequency of footbaths per week (OR=0.90; CI=081–0.99, for a 1-unit increase). The multivariable model included herd size, percentage of stalls containing fecal contamination, and presence of rubber in the alley to the milking parlor, and explained 44% of the variation of clinical lameness. Sev