Recorded as part of the Real Science Lecture Series, this webinar features Dr. Barry Bradford and Dr. Ghaffari. They move from the biochemistry of choline through two recent dose-response meta-analyses of rumen-protected choline (RPC) in transition cows.
Dr. Bradford opens by defining choline as a small, water-soluble nutrient often called a “quasi-vitamin.” It serves as a substrate for other molecules, including acetylcholine and phosphatidylcholine. The latter is important to cell membranes and lipoprotein assembly. Moreover, he notes choline was listed as indispensable for humans in 1998. (03:46 – 11:27)
From there, he examines one-carbon metabolism and epigenetics. He describes how choline, betaine, and methionine all feed into the pathway supplying S-adenosyl methionine. While this allows some metabolic flexibility, cow studies suggest these nutrients are not fully interchangeable. He also asks whether methyl donor supply around parturition could affect longer-term programming. (11:30 – 16:26)
Building on that foundation, he presents data on choline status across lactation. Cows sampled out to roughly 37 to 38 weeks show total plasma choline in later lactation well above the first three weeks. He also reviews a study in which a 300 g/d increase in choline chloride intake raised duodenal flow by only about 1.3 g/d. Additionally, he discusses an in vitro study finding 97% of supplied choline degraded at 15 minutes. (16:46 – 22:19)
With these findings in mind, Dr. Bradford walks through the requirements of rumen protection. A product must stay stable in the TMR, resist ruminal degradation, release choline post-ruminally, and deliver choline that is absorbed. Therefore, a stated feeding rate alone does not show what fraction reached the animal. He then argues recent evidence supports calling choline conditionally essential during transition. (22:20 – 26:41)
Dr. Ghaffari then takes over, charting choline publications from 1985 to 2026. He traces an early foundational phase and the first rumen-protected choline study in 1991. A transition-cow focus is seen through the 2000s and the first meta-analysis in 2010. This leads to a current precision era. Next, he details study selection. The 2025 production meta-analysis screened 429 records and yielded 30 studies. The 2026 metabolic analysis included 29 studies. (26:42 – 33:28)
Next, he explains the statistical framework. Because trials used no common dose structure, treatment effects within multi-dose experiments are correlated. A one-stage model accounts for that covariance and uses restricted cubic splines to estimate responses without forcing a straight line. Most active doses cluster between roughly 10 and 20 g/d choline chloride. (33:29 – 38:15)
The production results follow an intermediate-dose pattern. Prepartum dry matter intake showed no supported dose response. However, overall intake was significant, with a largest estimated response of +0.48 kg/d at roughly 13 to 14 g/d. This was significant from 12 to 19 g/d. Milk yield peaked at +1.29 kg/d at 13 g/d, significant from 11 to 20 g/d. In addition, 3.5% fat-corrected milk reached +2.19 kg/d, supported out to 21 g/d. (38:16 – 44:23)
He then connects these results back to biology. Once absorbed, choline enters phospholipid and one-carbon pathways in the liver. In the liver, phosphatidylcholine contributes to lipoprotein assembly and lipid export. As a result, he suggests RPC may support production through improved nutrient partitioning rather than intake alone. No dose response was found for milk fat, protein, or lactose percentage. However, milk fat yield was significant at +0.09 kg/d at 24 g/d, and milk protein yield from 13 to 20 g/d. (44:24 – 50:32)
Attention then turns to body reserves and blood metabolites. Body condition score showed no significant dose response, while overall body weight was significant at +12.5 kg at 11 g/d. Blood NEFA, BHB, and glucose showed no significant effect in either period. Dr. Ghaffari offers two hypotheses, explicitly not proven by the analysis. Blood concentration reflects net balance rather than flux, and rising intake and milk yield may partly offset one another. (50:33 – 54:44)
Remaining metabolic markers were outcome- and period-specific. Prepartum triglycerides decreased around 18 to 21 g/d. Cholesterol showed its greatest reduction at 7 g/d, and AST was significant only postpartum. Because these rest on few studies, he frames them as supportive evidence. He closes with four take-home messages. The rationale for choline during transition, the dependence of efficacy on delivery, the caution that nominal dose does not reflect delivered supply, and the endpoint-specific nature of responses concentrated at 10 to 20 g/d. (54:45 – 1:00:05)
Finally, a live Q&A session follows, covering pre-fresh versus post-fresh feeding, milk production level and choline requirements, extended feeding beyond transition, early-life programming, liver fat content, methionine interactions, beef cows, on-farm interpretation of pooled results, and study counts. (1:00:06 – End)
