Slide 1: What's happening to the older cows? Understanding the geriatric cow
What’s happening to the older cows? Understanding the geriatric cow Ian J Lean BVSc, DVSc, PhD, MANZCVS Scibus & University of Sydney Helen M Golder PhD Scibus & University of Sydney Jose EP Santos PhD University of Florida, USA Stephen E LeBlanc PhD University of Guelph, Canada Todd Duffield PhD Un
Slide 2: Why Bother?
Why Bother? • By changing the age structure we can produce > 16% more milk Age Structure of the herd • We have 8% more x-bred calves 0.4 or pure-bred heifers to sell 0.35 0.3 Proportion of herd • Can apply more selection pressure 0.25 0.2 • Decrease GHG intensity and total 0.15 0.1 0.05 • WHAT IS ST
Slide 3: What is the benefit to bother ratio?
What is the benefit to bother ratio? • Something may be a good opportunity but….. is it worth the effort? • Why is there a bother with getting that extra milk?
Slide 4: How did we ask the questions?
How did we ask the questions? • We got together with friends & colleagues to gather more than 36,000 cows of data from OZ, USA and Canada collected in carefully conducted studies • We wanted to look at the health, reproduction, body condition and metabolism of these cows to help direct our future st
Slide 5: What did our data base look like?
What did our data base look like? • 13 studies 32,783 Parity Country • We excluded non-Holstein 45 90 40 80 35 70 • TMR fed 28,675; Pasture 4,108 Percentage of cows Percentage of cows 30 60 25 50 • Key outcomes 20 40 15 • Hazard of not-being bred 10 30 20 • Time to pregnancy 5 10 • Pregnancy to firs
Slide 6: Statistical analysis
Statistical analysis • Survival analysis models were also used to evaluate time not bred over the first 120 DIM (HNBRED), and HPREG. Cox’s proportional Hazards conditions were met. • For pregnant in the lactation (OPAL) and Preg to 1st service LR used. • Similar models for disease and linear mixed m
Slide 7: Summary of prospective studies used in the survival database
Summary of 9 prospective studies used in the survival database Observation period Reference Study year Country Production system N cows Intervention (d) Duffield et al. (1999a) 1994 CA TMR 996 Monensin 95 Duffield et al. (2002) 1998 CA TMR 152 Monensin 120 LeBlanc et al. (2002) 1998 CA TMR 975 Vitam
Slide 8: What did our survival data look like?
What did our survival data look like? • Survival 30,780 Parity Country • We excluded non-Holstein 45 100 • The mean for all parities was 2.28 SD 1.46 40 90 80 35 Percentage of cows 70 Percentage of cows • Pasture 1,697 30 60 25 50 20 • Key outcomes 40 15 • All removals 30 10 • Deaths 20 5 10 • Sales
Slide 9: Kaplan-Meier survival curves for days to all removals by parity
Kaplan-Meier survival curves for days to all removals (deaths and culls) by parity
Slide 10: Parity, study year, and feeding system effects on time to death or sale
Parity, study year, and feeding system and interactions on time to death or being sold, time to sale only, or time to death only. Hazard ratio ± SE and P-values from models including parity, study year, and production system, and interaction between parity and production system. Parity 1 is referent
Slide 11: Summary of the studies used in the reproductive database
Summary of the 13 studies used in the reproductive database Period of Study Production Milk compo- observation Country N cows Intervention Reference year system nent data for pregnancy (d) Curtis (1997) 1992 Australia Pasture 237 N observational 368 Curtis and Lean (1998) protein & genetic Westwood
Slide 12: Hazard of not being bred
Hazard of not being bred • Kaplan-Meier survival curves for the days to removal before first insemination [i.e., the hazard of not being bred (HNBRED) or censoring for the hazard of being inseminated at least once] by parity. • The data are from 32,356 cows. • Parity ≥5 cows had 2.45 times greater H
Slide 13: Time to pregnant
Time to pregnant • Survival curves for days to pregnancy by parity • The data are from 32,969 cows in 13 studies • All parity groups differed in the daily probability of pregnancy which declined with parity • Parity 5+ v 1 HR 0.73
Slide 14: How is milk production involved?
How is milk production involved? • Association of milk yield (kg/d) at approximately 70 DIM centred on Group with relative predicted hazard of pregnancy • The data are from 20,071 cows • All parity groups differed in probability or hazard of pregnancy • Parity 1 cows had the greatest daily probabili
Slide 15: Milk protein% and relative predicted hazard of pregnancy
Milk protein% and relative predicted hazard of pregnancy • Association of milk protein percent centered on Group within herd at approximately 70 DIM with the relative predicted hazard of being pregnant. The data are from 8,761 cows in 7 studies. • Parity ≥5 cows had lesser daily probability of pregn
Slide 16: Associations of odds of pregnancy to first breeding with parity and milk production
Associations of odds of pregnancy to first breeding (PREG1) with parity, and milk production variables (approx. 70 DIM) controlling for parity, study year, and production system. Parity (P-value) Effect of milk Production Measure n Parity 1 is referent production variable system 2 3 4 ≥5 Linear Quad
Slide 17: The problem is overall pregnancy (OPAL), not so much at first service
The problem is overall pregnancy (OPAL) Not so much at first service • Milk yield (kg/d) at approximately 70 DIM Predicted coefficient for odds of pregnancy centred on Group and relative predicted coefficient for odds of pregnancy in a lactation • All parities differed, with parity 1 cows having the
Slide 18: Summary of responses for milk production and reproductive outcomes
Summary of responses for milk production, milk fat percentage and yield, milk protein percentage and yield, and yield of milk solids with reproductive outcomes. ‒ L indicates a negative linear association and Q a quadratic association. Hazard of not Time to Pregnancy to Pregnancy in a Milk measures
Slide 19: Summary of reproduction?
Summary of reproduction? • Reproductive failure was the major source of removal in this study • In general, increased parity reduces reproductive performance • However, very similar performance at first insemination • It is a failure to be mated and ultimately not going in calf • BOTH low production
Slide 20: Are there any problems with increasing longevity?
40 OR parity 5+ 35 Parity 1 = 2.5 parity 1 Average 19% Parity 2 Incidence of disorder (%) 30 With OR greatest Parity 3 OR parity 5+ For parity 1 Parity 4 = 5.6 parity 1 25 OR parity 5+ Parity ≥5 20 = 2.3 parity 1 OR parity 5+ 15 = 4.9 parity 1 10 5 0 Milk fever Dystocia Retained Metritis Endometriti
Slide 21: Disorder odds by parity
Parity1 (P-value) Disorder n 2 3 4 ≥5 Clinical 3.52 ± 0.74bde 8.61 ± 1.78bc 20.02 ± 3.94bcd 15,793 ‒ hypocalcemia3 (<0.001) (<0.001) (<0.001) 1.16 ± 0.05acde 1.69 ± 0.08abde 1.91 ± 0.11abce 2.46 ± 0.15 abcd Mastitis 27,857 (0.001) (<0.001) (<0.001) (<0.001) 1.55 ± 0.10acde 3.09 ± 0.21abde 4.32 ± 0.3
Slide 22: Glucose and parity
Glucose and parity: Difference from group centered mean pre-calving, at calving, immediately after calving and at peak Difference mmol/L from Group Centered Glucose 0.6 0.4 0.2 Mean 0 -0.2 -0.4 -0.6 1 2 3 4 ≥5 Parity Pre-calving (-3 to -1 DIM) Calving Date Post-Calving (1-3 DIM) Peak Milk
Slide 23: NEFA and parity
NEFA and parity: Difference from group centered mean pre- calving, at calving, immediately after calving and at peak NEFA Difference mmol/L from Group Centered Mean 0.2 0.15 0.1 0.05 0 -0.05 -0.1 -0.15 -0.2 1 2 3 4 ≥5 Parity Pre-calving (-3 to -1 DIM) Calving Date Post-Calving (1-3 DIM) Peak Milk
Slide 24: BHB and parity
BHB and parity: Difference from group centered mean pre- calving, at calving, immediately after calving and at peak 3-Hydroxybutyrate 0.25 Difference mmol/L from Group Centered 0.2 0.15 0.1 0.05 0 Mean -0.05 -0.1 -0.15 -0.2 1 2 3 4 ≥5 Parity Pre-calving (-3 to -1 DIM) Calving Date Post-Calving (1-3
Slide 25: Cholesterol and parity
Cholesterol and parity: Difference from group centered mean pre- calving, immediately after calving and at peak Cholesterol 0.5 Difference mmol/L from Group Centered 0.4 0.3 0.2 0.1 0 Mean -0.1 -0.2 -0.3 -0.4 1 2 3 4 ≥5 Parity Pre-calving (-3 to -1 DIM) Post-Calving (1-3 DIM) Peak Milk
Slide 26: Calcium and parity
BHB and parity: Difference from group centered mean pre- calving, at calving, immediately after calving and at peak Calcium 0.2 Difference mmol/L from Group Centered 0.15 0.1 0.05 0 Mean -0.05 -0.1 -0.15 -0.2 1 2 3 4 ≥5 Parity Pre-calving (-3 to -1 DIM) Calving Date Post-Calving (1-3 DIM) Peak Milk
Slide 27: Summary of longevity & Metabolism?
Summary of longevity & Metabolism? • There are differences in survival with older cows ie 5+ parity markedly more likely to die. They are at greater risk for removal overall. • There are marked differences in concentrations of metabolites for cows of 5+ parity that are reflected in differences in di
Slide 28: BCS across Parity
BCS across Parity BCS/weigh before < 30 DIM and in-lactation 7 to 100 DIM: Difference from Group centered mean Pre-Calving BCS Peak Milk BCS Change in BCS
Slide 29: What might be causing this? Cows gain weight but lose BCS
What might be causing this? Cows gain weight but lose BCS • Within parity, expect 25% for each BCS/BW category • Parity 1: 61% High BCS/Low BW • Parity ≥5: 56% Low BCS/High BW
Slide 30: What might be causing this? Cows gain weight but lose BCS
What might be causing this? Cows gain weight but lose BCS Proportion of cows in each category of BCS and BW Proportion of cows in each category of BCS and BW Holstein Other Breeds 70 Parity 1 70 Parity 1 Parity 2 60 Parity 2 60 Proportion of cows (%) Proportion of cows (%) Parity 3 Parity 3 50 50 Pa
Slide 31: Where to now? Define and improve
Where to now? Define and improve • We have strong evidence of impaired metabolism with increased parity (another presentation). • In order to fix it, we need to better define the causes BUT we have many solutions available – ration balancing, better protein nutrition, fat nutrition, some reproductiv
Slide 32: Thanks
Thanks • Collaborators • Past students who contributed data sets • Colleagues in Dairy UP • Corporate and public good supporters of the research included in the original studies
Slide 33: Statistical analysis - BCS
Statistical analysis - BCS • All outcomes (pre-calving BCS, peak-milk BCS, BCS change and peak-milk BW) were all centered around study Group mean to help control for study year, genetic differences and other unknown confounding variables between studies. • Ordinary least-squares linear regressions p