Cross-sectional study of the effect of diet composition on plasma folate and vitamin B12 concentrations in Holstein cows in the United States and Canada

Abstract

The objective of this cross-sectional study was to assess the variability of plasma folate and vitamin B12 concentrations in lactating Holstein cows across the United States and Canada. We also evaluated the effect of diet composition and cow characteristics on folate and vitamin B12 plasma vitamin concentrations. A total of 22 and 24 US and Canadian dairy herds were enrolled, totaling 427 and 476 cows at 10 to 197 days in milk across all US and Canadian herds, respectively. Blood samples were taken to analyze plasma folate and vitamin B12 concentrations, and ingredients of the diet were collected to determine nutrient composition. To reduce the number of interdependent variables in the analysis of the association of diet composition with plasma vitamin concentrations, we conducted a principal component analysis. Plasma folate concentrations were lower for US cows [13.4 ng/mL, 95% confidence interval (CI): 12.7–14.2] than for Canadian cows (14.5 ng/mL, 95% CI: 13.7–15.2), and the opposite was observed for plasma vitamin B12 concentrations (US 206 pg/mL, 95% CI: 192–221; Canada 170 pg/mL, 95% CI: 159–181). The highest plasma concentrations of both vitamins were observed in the Northwest region of the United States (Oregon and Washington). Cows in California had the lowest plasma folate concentrations, and cows in Québec and New York State had the lowest plasma vitamin B12 concentrations. Plasma folate concentrations were higher for multiparous than for primiparous cows and plasma vitamin B12 concentrations progressively increased from parity 1 to 3 and higher. For both studied vitamins, plasma concentrations were lower at 0 to 55 than at 56 to 200 days in milk. Of 3 principal components, the one associated with dietary carbohydrates was significantly correlated with plasma folate and vitamin B12 concentrations. Indeed, plasma folate concentrations decreased with dietary fiber concentrations (i.e., neutral and acid detergent fibers and lignin) and increased with dietary nonfiber carbohydrate concentrations. We obtained the opposite results for plasma vitamin B12 concentrations. Both multivariable models explained 41% (pseudo-R2) of the variation in plasma folate and vitamin B12 concentrations. Information gathered in this study is the first step toward determining sources of variation in plasma folate and vitamin B12 concentrations, as well as the vitamin status of cows.