Grad Student Competition
Effects of Propylene Glycol on Milk Yield in Multiparous Cows based on Early Lactation Rumination and Eating Behavior: A Randomized Controlled Trial
| Date/Time: |
9/12/2025
08:45
|
| Author: |
Sara Davis |
| Clinic: |
University of Minnesota |
| City, State, ZIP: |
Cokato, MN 55321 |
Sara Davis, DVM
1
;
Marina Dorella, DVM
1
;
Cainan Florentino, DVM
1
;
Megan Ruch, BS
1
;
Elise Shepley, MSc, PhD
1
;
Luciano Caixeta, DVM, PhD
1
;
1Department of Veterinary Population Medicine, University of Minnesota, Saint Paul, MN 55321
Introduction:
In a shrinking workforce with increasing demands for labor efficiency, it's essential to target treatments to cows most likely to benefit. Although propylene glycol (PG) is commonly used to treat hyperketonemia (HYK), prior research has shown conflicting results. Early lactation rumination and eating behavior may help identify cows in need of treatment. This study evaluated the effect of PG treatment on milk yield in cows with varying levels of early postpartum rumination and eating time.
Materials and methods:
In a randomized trial on a Minnesota dairy, multiparous cows were screened for HYK at 3 ± 1 and 7 ± 2 DIM using a hand-held BHB meter. Cows with BHB ≥ 1.2 mmol/L were classified as HYK and randomized to PG (300 g orally for 3 days; n = 250) or no treatment (NoPG; n = 226). Non-HYK cows (HYK–; n = 2,180) served as a comparison group. Farm staff treated cows for clinical HYK as needed, with all treatments in the first 30 DIM recorded for HYK– and NoPG cows, as well as for PG cows receiving additional treatment after enrollment. Cow-level and production data were sourced from herd management software, while rumination and eating time were captured via wearable sensors and expressed as minutes per day (converted from total daily proportion). All cows were followed until 150 DIM or herd removal.
For statistical analysis, rumination and eating time were averaged over the first 7 DIM and used to categorize cows into three groups for each behavior outcome (lower 25%, middle 50%, upper 25%). Milk yield over the first 22 weeks was analyzed via a linear mixed effects model incorporating fixed effects for treatment group (PG, NoPG, HYK–), behavioral group, and their interaction, along with covariates including parity, prior 305 ME, clinical ketosis treatment by the farm, baseline rumination or eating behavior (–14 to –7 DIM), BHB at diagnosis (or equivalent time period for HYK– cows), and DIM at BHB measurement. Week in lactation served as a repeated measure with an autoregressive (AR1) correlation, and cow ID was treated as a random effect.
Results:
Descriptively, compared to HYK– cows, PG and NoPG cows were more likely in ≥3rd lactation (75%, 73% vs. 49%), and had lower average rumination (361, 348 vs. 438 min/day) and eating times (233, 225 vs. 281 min/day). They were also more often in the lowest rumination (46%, 53% vs. 21%) and eating quartiles (42%, 47% vs. 22%) than HYK− cows.
Across all rumination quartiles, PG cows had the lowest milk yield. In the middle rumination group, mean yields were 33.0 kg/day (95% CI: 27.0, 39.0) for PG, 42.8 kg/day (95% CI: 36.0, 49.6) for NoPG, and 39.3 kg/day (95% CI: 36.7, 42.0) for HYK– cows. The difference between NoPG and PG was statistically significant (P < 0.01). In the low rumination group, yields were 29.4 kg/day (95% CI: 23.3, 35.4) for PG, 35.1 kg/day (95% CI: 28.8, 41.3) for NoPG, and 35.3 kg/day (95% CI: 32.1, 38.6) for HYK–. The NoPG vs. PG difference was 5.7 kg/day (P = 0.21); HYK– vs.PG was 6.0 kg/day (P = 0.06), suggesting PG cows consistently underperformed.
Similar trends occurred when stratifying by eating time. In the low eating group, PG, NoPG, and HYK– cows produced 28.7 kg/day (95% CI: 23.8, 33.6), 36.9 kg/day (95% CI: 32.0, 41.7), and 34.9 kg/day (95% CI: 32.5, 37.4), respectively, with NoPG vs. PG significant (P < 0.001).
Significance:
Cows treated with PG consistently produced less milk over the first 22 weeks postpartum compared to NoPG and HYK– cows, regardless of rumination or eating activity. These results suggest PG treatment provided no benefit and, in many cases, reduced milk yield compared to untreated cows. Our findings do not support using PG to improve early lactation milk yield in hyperketonemic cows under these conditions. Further research is needed to explore underlying mechanisms and identify better treatment or management strategies.
