Residual feed intake in peripartal dairy cows is associated with differences in milk fat yield, ruminal bacteria, biopolymer hydrolyzing enzymes, and circulating biomarkers of immunometabolism

A. A. Elolimy, Y. Liang, K. Wilachai, A. S. Alharthi, P. Paengkoum, E. Trevisi, J. J. Loor

Research output: Contribution to journalArticlepeer-review


Residual feed intake (RFI) measures feed efficiency independent of milk production level, and is typically calculated using data past peak lactation. In the current study, we retrospectively classified multiparous Holstein cows (n = 320) from 5 of our published studies into most feed-efficient (M-eff) or least feed-efficient (L-eff) groups using performance data collected during the peripartal period. Objectives were to assess differences in profiles of plasma biomarkers of immunometabolism, relative abundance of key ruminal bacteria, and activities of digestive enzymes in ruminal digesta between M-eff and L-eff cows. Individual data from cows with ad libitum access to a total mixed ration from d −28 to d +28 relative to calving were used. A linear regression model including dry matter intake (DMI), energy-corrected milk (ECM), changes in body weight (BW), and metabolic BW was used to classify cows based on RFI divergence into L-eff (n = 158) and M-eff (n = 162). Plasma collected from the coccygeal vessel at various times around parturition (L-eff = 60 cows; M-eff = 47 cows) was used for analyses of 30 biomarkers of immunometabolism. Ruminal digesta collected via esophageal tube (L-eff = 19 cows; M-eff = 29 cows) was used for DNA extraction and assessment of relative abundance (%) of 17 major bacteria using real-time PCR, as well as activity of cellulase, amylase, xylanase, and protease. The UNIVARIATE procedure of SAS 9.4 (SAS Institute Inc.) was used for analyses of RFI coefficients. The MIXED procedure of SAS was used for repeated measures analysis of performance, milk yield and composition, plasma immunometabolic biomarkers, ruminal bacteria, and enzyme activities. The M-eff cows consumed less DMI during the peripartal period compared with L-eff cows. In the larger cohort of cows, despite greater overall BW for M-eff cows especially in the prepartum (788 vs. 764 kg), no difference in body condition score was detected due to RFI or the interaction of RFI × time. Milk fat content (4.14 vs. 3.75 ± 0.06%) and milk fat yield (1.75 vs. 1.62 ± 0.04 kg) were greater in M-eff cows. Although cumulative ECM yield did not differ due to RFI (1,138 vs. 1,091 ± 21 kg), an RFI × time interaction due to greater ECM yield was found in M-eff cows. Among plasma biomarkers studied, concentrations of nonesterified fatty acids, β-hydroxybutyrate, bilirubin, ceruloplasmin, haptoglobin, myeloperoxidase, and reactive oxygen metabolites were overall greater, and glucose, paraoxonase, and IL-6 were lower in M-eff compared with L-eff cows. Among bacteria studied, abundance of Ruminobacter amylophilus and Prevotella ruminicola were more than 2-fold greater in M-eff cows. Despite lower ruminal activity of amylase in M-eff cows in the prepartum, regardless of RFI, we observed a marked linear increase after calving in amylase, cellulase, and xylanase activities. Protease activity did not differ due to RFI, time, or RFI × time. Despite greater concentrations of biomarkers reflective of negative energy balance and inflammation, higher feed efficiency measured as RFI in peripartal dairy cows might be associated with shifts in ruminal bacteria and amylase enzyme activity. Further studies could help address such factors, including the roles of the liver and the mammary gland.

Original languageEnglish (US)
Pages (from-to)6654-6669
Number of pages16
JournalJournal of Dairy Science
Issue number8
StatePublished - Aug 2022


  • feed efficiency
  • lactation
  • nutrition
  • ruminal microbiota

ASJC Scopus subject areas

  • Food Science
  • Genetics
  • Animal Science and Zoology


Dive into the research topics of 'Residual feed intake in peripartal dairy cows is associated with differences in milk fat yield, ruminal bacteria, biopolymer hydrolyzing enzymes, and circulating biomarkers of immunometabolism'. Together they form a unique fingerprint.

Cite this