Negative energy balance and Ketosis: Consequences and monitoring in. cows. Collaborators: on the transition cow side

Transcription

1 Negative energy balance and Ketosis: Consequences and monitoring in transition dairy cows Daryl Nydam, DVM, PhD Cornell University Learning Objectives: 1. Basic transition cow metabolism 2. Understanding cow-level metabolite indicators for downstream performance 3. Understanding herd-level metabolite indicators for downstream performance 4. Develop rational sampling plans to detect cows and herds at risk of poor performance 5. Appreciate some interventions for cows and herds at risk of poor performance due to negative energy balance Stars of the show Collaborators: on the transition cow side Tom Overton, An Sci Yves Boisclair, An Sci Tracy Stokol, PMDS Julia Flaminio, DCS Rob Gilbert, DCS Joe Wakshlag, DCS Gary Oetzel, UW-Madison NEB and SCK in transition dairy cows Rapid change in cow s metabolic demands 2x Energy requirements Energy (NE L ) Requirements 2 Days Before Versus 2 Days After Calving 725-kg Cow 575-kg Heifer Function Pre Post Pre Post Maintenance Pregnancy Growth Milk production Total (Mcal) Calculated from NRC (2001). Assumes milk production of 25 kg/d for cow and 20 kg/d for heifer, each containing 4% fat. Courtesy of J. K. Drackley 1

2 Epi Body fat Fat Insulin CO 2 Propionate Liver Mitochondrial Fuel use Non-esterified Fatty Acids () In blood, come from adipose tissue Mostly from mobilization of triglyceride A little from dietary fats Mammary Gland Milk Fat VLDL Ketone Bodies TG Fat Measure of Negative Energy Balance Related to post-partum disease, performance, and reproduction? Stolen from Drackley via Overton (Subclinical) Ketosis (SCK) Ketosis is the elevation of ketone bodies: acetoacetic acid Acetone (not metabolized, expelled in lungs) β-hydroxybutyric acid (BHB) Most stable in blood Punch line #1 Transition dairy cows have a huge challenge All cows visit NEB in this period Can measure extent of it with and BHB Results from negative energy balance (NEB) in early lactation or insulin resistance starting in dry period May be secondary to diseases causing decreased feed intake The questions What are the impacts of NEB and SCK in modern dairy cows? All cows visit NEB in early lactation How do we know when a cow is at risk for poor performance? How do we know when a herd is at risk for poor performance? 2

3 Objectives: Cow-level 1. Evaluate metabolites and their association with clinical disease, reproduction, and milk production in larger herds in the northeastern USA 2. Establish cow-level critical thresholds for and BHB Materials and methods Prospective cohort study Cross-sectional sampling 1 visit Convenience sample of herds > 250 milking cows Free-stall TMR-fed DHIA/ DC 305 herds in the NE USA Convenience sample of transition animals Healthy! 14-2 days pre-partum 3-14 days post-partum 1/3 heifers 2/3 cows 15 cows/herd Sampling Day COW NEB and SCK in transition dairy cows Data Collection Predictors:, BHBA (other metabolites) Herd, BCS, DCC, DIM, LACT, season, diet, stocking density, pen moves, time of sampling, etc Outcomes: Clinical disease Metritis Retained Fetal Membranes (RP) Ketosis (CK) Displaced Abomasum (DA) Reproduction ME 305 milk production Removal from the herd NEB and SCK in transition dairy cows Materials and Methods DC 305 DairyOne database Lab Analysis Colorimetric Wet Chemistry (Hitachi 917, Roche) Data Analysis (SAS V 9.1) Univariate Analysis PROC Univariate, PROC Freq, PROC Lifetest Multivariable Analysis and ROC curves PROC Logistic, PROC Genmod, PROC Phreg, PROC mixed, MedCalc 100 herds Results Mean= 840 cows (range: ) 2758 animals Heifers: 991 Cows: 1767 Lactational Incidence (%) Disease Outcomes DA Ketosis RP/Metritis Any of Median DIM /5 at Diagnosis Median ME305 Milk Production at 120 DIM: 27,000 lbs Median 21-day Pregnancy Rate:

4 How to figure out what /BHB values are important? Receiver Operator Characteristic (ROC) curve Plots True Positive Rate v. False Positive Rate for all observed values in the data set against outcome of choice Point with greatest combo of Sensitivity and Specificity could be useful Area Under the Curve (AUC) tells how predictive the variable is of the outcome; the larger the better.5 = coin flip.5 to.7 = accurate test.71 to.9 = very accurate test >.91 = highly accurate and rare ROC Curve POST-partum : 0.72 (meq/l) DA Sensitivity Sensitivity: 80.5 Specificity: 72.7 Criterion : > Specificity AUC:0.81 Highly accurate Impact on disease outcome? > 0.3 meq/l in animals pre-partum Disease Relative Risk P-value DA CK RP/Met 2.2 < Any of < Impact on disease outcome? BHB > 10 mg/dl in animals post-partum Disease Relative Risk P-value DA 6.9 < CK 4.9 < Met Any of < Impact on disease outcome? > 0.6 meq/l in animals post-partum Disease Relative Risk P-value DA 9.7 < CK 5.0 < RP/Met 17 <0.008 Any of < Punch line #2 Predicting any Dz: Sample > animals Risk Factor Threshold Risk Ratio AUC Pre-partum (meq/l) Post-partum BHB (mg/dl) Post-partum (meq/l) Accurate Highly accurate Highly accurate 4

5 Impact on Reproduction? Outcome: Conception within 70 days post-voluntary waiting period 95% of herds using ovulation synchronization protocol Equal time at risk for all herds Optimal calving interval Covariates: BCS, herd, season, parity, milk production Reproduction: PRE-partum - time to pregnancy t not pregnant Percent meq/l < 0.27 meq/l Days to conception from voluntary waiting period NEB effect on Reproduction Kaplan-Meier curves of time to pregnancy of 1318 cows and heifers with 0.27 meq/l or < 0.27 meq/l measured in serum 14-2 days pre-partum. Results- Reproduction Animals sampled POST-partum Cox proportional hazard models Variable Post-partum >0.72 meq/l P-value Hazard Ratio Parity > Covariates: BCS, season, ME 305, BHB: p-values > 0.2 Results- Reproduction Animals sampled POST-partum Cox proportional hazard models Variable P-value Hazard Ratio BHB > mg/dL Parity > Covariates: BCS, season, ME 305: p-values > 0.2 5

6 Impact on Milk production? ME 305 evaluated at 120 DIM ME 305 milk can be compared between heifers and cows 120 DIM ME 305 most accurate measure of milk production in our study Covariates: BCS, season, herd, parity, and interaction between parity and metabolite NEB effect on Milk Production Results- Production Animals sampled PRE-Partum; Mixed effects models Variable P- value Difference in ME milk yield (kg) Pre-partum < > 0.33mEq/L Parity > Results- Production COWS sampled POST-Partum Partum Mixed effects models Variable P- Difference in ME value milk yield (kg) BHB > 10mg/dL Covariates: BCS, season: p-values > 0.2 Covariates: BCS, season, interaction: p-values > 0.4 Results- Production COWS sampled POST-Partum Mixed effects models Variable P- Difference in ME value milk yield (kg) Post-partum > 0.72 meq/l Covariates: BCS, season, BHB: p-values > 0.3 Punch line #3 Relatively small amount of NEB or SCK largely impacts reproductive performance and milk production Current work suggests this may be due to immune compromise and subsequent endometritis, as well as ovarian function ~14-18% less likely to get pregnant ~1000 lbs less ME305 milk Except heifers.. 6

7 NEB and SCK in transition dairy cows How much opportunity is there to better manage a NEB and SCK in transition animals costly condition? 1472 pre-partum animals 45% (245/540) heifers 0.3 meq/l 26% (246/932) cows 0.3 meq/l NEB and SCK in transition animals 1315 post-partum animals 25% (131/517) heifers 0.7 meq/l NEB and SCK in transition dairy cows Results Unrecognized Clinical ketosis at 30 mg/dl: 3 % (95% CI: %) 33% (267/798) cows 0.7 meq/l 15% (77/517) heifers BHB 10 mg/dl 27% (214/798) cows BHB 10 mg/dl Postpartum and Culling Risk Cull < 30 DIM 0.70 meq/l < 0.70 meq/l Total Yes No Postpartum and Culling Risk Mixed effects model with herd as a random effect Proc Genmod (SAS v. 9.1) Risk ratio = % CI = P-value: 0.02 Total Risk ratio: 3.9 (95% CI: 1.6 9) Chi-Square p-value:

8 Postpartum BHB and Culling Risk Cull < 30 DIM 0.70 meq/l < 0.70 meq/l Total Yes No Postpartum BHB and Culling Risk Mixed effects model with herd as random effect. Proc Genmod (SAS v. 9.1) Risk ratio = % CI = 2 18 P-value: Total Risk ratio: 6.2 (95% CI: ) Chi-Square p-value: < Punch Line #4 Lots of cow-level high and BHB Opportunity for cow-level management to decrease clinical disease, increase milk production, and improve reproductive performance High and BHB increase risk of culling in the post-partum period Herd alarm levels Herd alarm levels have not been well defined Oetzel 2004, VCNA Proportion of sampled animals with elevated metabolite levels? Herd level effects on health, reproduction and production Objectives: Herd-level 1. Determine the herd alarm level for the proportion of sampled animals with elevated metabolites resulting in herdlevel effects 2. Describe the herd-level prevalence of elevated metabolite concentrations. 8

9 Materials and Methods 1. Diseases: percent of cases of DA, CK (within 30 DIM) in animals sampled 2. Reproduction: herd pregnancy rate calculated over 2, 21-day yperiods post-herd-voluntary waiting period 3. Production: Average ME 305 milk (evaluated at 120 DIM) from animals sampled Heifers and cows were grouped in final analysis if outcomes between groups were similar Herd alarm level: Disease Metabolite Proportion Concentration Change in % of DZ P- value Pre-partum 15 % 0.27 meq/l Post-partum 15 % 0.70 meq/l BHB 15 % 12 mg/dl Herd alarm level: 21 d herd PR Metabolite Proportion Concentration Change in % of PR Pre-partum Post-partum P-value 15 % 0.27 meq/l % 0.70 meq/l BHB 15 % 12 mg/dl Herd alarm level: Milk Production Metabolite Proportion Concentration Change in ME 305 (kgs) Pre-partum P-value 15 % 0.27 meq/l Postpartum 15 % 070mEq/L < COWS BHB COWS 15 % 10 mg/dl -358 <0.01 Postpartum HEIFERS 15 % 0.60 meq/l BHB HEIFERS 20 % 12 mg/dl -534 <0.01 Metabolite level PRE-Partum > 0.3 meq/l POST-Partum > 0.6 a -0.7 b meq/l BHB > 10 a -12 b* mg/dl Impact / Punch Line #5 Herd Alarm Effect 15% -1.2 PR +1.4% DZ -240 kg ME305 milk POST P t 15% PR a 15% *20% *15% of 15 = 2-3 animals +1.3% DZ b Heifers:-290kg,Cows:-577kg -0.7 PR b +1.2% DZ b Heifers: -535kg*, Cows:-332kg a What s the herd-level opportunity? 9

10 74% of herds had >15% of animals sampled PRE-partum with > 0.30 meq/l Animals sampled pre-partum 65% of herds had > 15% of animals sampled POST-partum with > 0.70 meq/l Animals sampled post-partum % of herds < 15% >15 <25% > 25 - < 35% > 35% % of herds <15% > 15 - < 25% > 25 - <35% > 35% % of animals/herd with > 0.30 meq/l % of animals with > 0.70mEq/L 42% of herds had > 15% of animals sampled POST-partum with BHB > 12 mg/dl % of herds Animals sampled post-partum 21 <15% > 15 - < 25% > 25 - <35% > 35% % of animals with BHB > 12mg/dL 13 8 Punch Line #6 Lots of opportunity for better management!! Herds and cows not selected because of problems Unrecognized clinical disease Cost of DA, metritis/rp, ketosis etc. Reproductive Performance suffers Lost milk Two Applications of Ketosis Testing Identifying individual ketotic cows cowside test for treatment decisions how intensive should we be? Identifying herds with ketosis problems herd-level test for management decisions Trouble-shooting On-going monitoring When is Individual Sick Cow Testing Warranted? Reasons to test a cow: Not eating well Looks dull (eyes, ears) Atypical behavior Low milk yield Other disease problem Only in early lactation 10

11 When is Individual Cow Testing Warranted? Blanket testing possibilities: M -W -F or one day a week or when exiting the fresh pen or... Blanket testing would be warranted if early treatment was beneficial How to Determine Prevalence? Three fluids can be sampled urine milk blood Sweet smell of breath acetone, other compounds only about 50% sensitive Urine Ketones Ketostix test strip about $0.20 each have to stimulate urination dip and read within 10 seconds acetoacetate causes nitroprusside reagent to turn purple 40 to 60% will urinate Urine Ketones catheters or vaginal exams are impractical touching strip to vaginal walls is inaccurate Need a plan for cows that do not urinate Milk Ketone Tests Powders or tablets nitroprusside reaction 35% sensitive 98% specific Not recommended as sole test Milk Ketone Tests Ketotest test strip $2.00 per test dip and read 60 seconds later use 100 umol/l cut point 83% sensitive 82% specific 11

12 Cow-side tests for SCK (compared to 14 mg/dl serum BHB) Milk Keto-test 100 mmol/l Sen: 83% Spec: 82% 200 mmol/l Sen: 54% Spec: 94% Powder Lacks good sensitivity Urine Ketostix Read in 5 seconds Sen: 75% Spec: 96% Acetest tablet Lacks good specificity Precision XTRA 1 drop of blood for meter, cow-side answer in 10s Compared serum BHB Sensitivity: % Specificity: % Heuweiser JDS 09; Oetzel AABP 08; Burke MSU 07 ~ $1.20/test Laboratory Analysis Use a lab with validated tests Red tops tubes preferred for both BHB and Keep sample chilled and separate serum ASAP Arrive at lab with 1-2 days Can freeze serum for up to 1 month Can get away with separating serum <24 hours and then sample in fridge for 3 days At Cornell, and BHB are $11.00/test Statistical Sampling: Sample: A partial selection of animals from a larger group (herd) from which blood, tissues, etc are collected for testing. A judgment will be made about the larger group based on the findings from the sample. Tells us how lazy and cheap we can be: Not enough time to see all the cows Note enough money to test all the cows BUT, we re trying to make money by being correct at least some of the time So we learn how to sample! 12

13 How confident do you need to be to make management decisions? 99%? 95%? 90%? 75%? The more confident you want to be the more cows you have to sample! What proportion of my cows have Y? Need to know: Desired confidence about your fraction The error you are willing to accept Prevalence above which you care Population Size BHBA and suggestions 75% confident that sample proportion represents true population proportion If >15% of cows are in Negative Energy Balance, then it s a herd problem Want to estimate true proportion +/- 10% Target population has some effect, we ll say 35 cows at risk for NEB If you want to be 90% confident, the sample size becomes 15 cows Punch line #7 Sample size: >12 cows at risk, 15 is better Cows at risk to sample Pre-partum: 14 to 2 days before calving Post-partum: 3 to 14 DIM; a better predictor, BHB cheaper and easier to handle Sample to take Serum (red top tubes) Don t shake, keep cool What to do with sample? BHB: Precision Extra : Lab What to do with results Interpret % above cut-point Currently suggesting herd-level problem if > ~15-20% above cutpoint for and BHB Comparison between individual and pooled samples of and BHBA in transition dairy cows for herd energy evaluation Paula A. Ospina 1, DVM, MPH, PhD candidate; Daryl V. Nydam 2, DVM, PhD; Tracy Stokol 2, BVSc, PhD; Tom R. Overton 1, PhD; 1 Cornell University, Dept. of Animal Science 2 Cornell University, Dept. of Pop. Med. and Diag. Science 13

14 Materials and methods Individual tests: Evaluated and BHBA in individual samples from 12 herds Pooled sample: 200 μl aliquots from each individual sample from the 12 herds Mc Nemar s: Assess Bias Kappa Do the tests agree? Sensitivity and Specificity Pooled vs. proportion (individual tests) Results Metabolite Pre-partum Post-partum McNemar s p-value Kappa Kappa Interpretation Slight agreement Fair agreement BHB Slight agreement Results Metabolite Se 95% CI for Sensitivity Spe 95% CI for Specificity The next questions What do we do about these cows with SCK? Pre-partum 55% 24-83% % What do we do about herds with SCK? Postpartum 50% 19-81% % BHB 30% 7-65% % Important stuff to not forget: Access to feed/crowding/social dynamics: 30 of linear feed space or 80% of headlocks/stalls Few group/pen changes as possible Separate parity=1 from parity>1 Energy density of dry cow rations (esp. far off) Mcal/lb NEL 12-16% starch 40-50% forage NDF Mcal/d far-off and close-up Important stuff not to forget: Feeding rumensin? Feeding rumen protected choline? Heat abatement! Feeding other products that affect metabolism and/or liver physiology 14

15 Effect of Propylene Glycol on Subclinical Ketosis Resolution of ketosis and milk yield J. A. A. McArt, D. V. Nydam, P. A. Ospina, and G. R. Oetzel Propylene Glycol (PG) Two modes of action: Increased supply of propionate Glucogenic Insulin resistance Decrease in glucose demand by peripheral tissues Purpose of Study To determine the effect of oral PG administration in cows diagnosed with SCK in early lactation: Ketosis resolution Clinical ketosis prevention Milk yield Displaced Abomasum Herd Removal Precision Xtra TM Ketone System Requires 1.5 µl of whole blood Result in 10 seconds $1.30 per test Compared to serum BHBA Sensitivity: % Specificity: % Iwersen et al., 2009; Konkol et al., AABP 2009 Study Herds Two free-stall dairies in NY Farm A: 92 lbs milk (1900 cows) Farm B: 92 lbs milk (1825 cows) Two free-stall dairies in WI Farm C: 86 lbs milk (2800 cows) Farm D: 77 lbs milk (4100 cows) 15

16 Study Design Data collected from May - August, 2010 Cows tested 6 times between 3 16 DIM M, W, F testing Subclinical ketosis mm/l Sequential randomization to treatment group Clinical ketosis 3.0 mm/l Treated according to farm protocol Subclinical ketosis mm/l Treatment group 300 ml PG Once daily drench until either: Test < 1.2 mm/l 17 DIM Control = no PG Statistical Analysis Semiparametric proportional hazards model Time to ketosis resolution Time to development of clinical ketosis Repeated measures ANOVA Milk weights Stratified by herd Cow Enrollment 2115 cows enrolled Removed from analysis PDCC < 260 Died/sold Farm ketosis treatment Improper identification Inappropriate p randomization Fewer than 5 tests 1717 cows in final analysis 976 non-ketotic cows 741 ketotic cows randomized to treatment 372 PG cows 369 control cows Subclinical Ketosis Incidence Prevalence of SCK by DIM Herd Location A NY B NY C WI D WI All --- Cows SCK (%) BHBA mm/l (%) Days in Milk 16

17 9 mm/l (%) BHBA Incidence of SCK by DIM Days in Milk Cows with SCK (%) Resolution of Ketosis Treatment Group: Control PG PG treated cows (n=372) were 1.5 X more likely to resolve their ketosis than control cows (n = 369) 95% CI = 1.3 to 1.8; P < Time (days) Cows Progressing g to CK (%) Progression to Clinical Ketosis PG treated cows (n = 372) were 0.5 times as likely to develop CK as control cows (n = 369) 95% CI= 0.3 to 0.9; P = Treatment Group: Control PG Time (days) Milk Yield Daily milk weights for herds A, B, and D 622 cows with at least one test mm/l Overall milk yield response (first 30 DIM): 0.7 more lbs/cow/day for ketotic cows treated with PG (P < 0.001) Milk response by herd Herd A = 3.0 lbs milk (P < 0.001) Herd B = 3.5 lbs milk (P < 0.001) Herd D = 0.2 lbs milk (P = 0.70) Ketosis, DA, and Removal Incidence Herd SCK (%) DA (%) Early Removal (%) Farm A 40.4 (n = 143) 2.0 (n = 7) 2.3 (n = 8) Farm B 26.4 (n = 106) 2.0 (n = 8) 3.0 (n = 12) Farm C 40.9 (n = 119) 1.7 (n = 5) 7.2 (n = 21) Farm D 55.7 (n = 373) 3.3 (n = 22) 8.8 (n = 59) DA Incidence by Treatment Herd A B C D DA Control 4.3% 9.3% 5.2% 7.1% PG 4.1% 3.8% 3.3% 4.2% Control cows were 1.6 times more likely to develop a DA than PG treated cows 95% CI = 1.3 to 2.0, P <

18 Removal Incidence by Treatment Herd A B C D Removal Control 5.7% 3.7% 3.4% 6.5% PG 1.4% 0.0% 1.6% 4.2% Control cows were 2.1 times more likely to be removed from the herd than PG treated cows Conclusions Oral PG helps! Speeds resolution of ketosis Prevents progression to severe ketosis Increases early lactation milk yield in some herds Reduces risk of DA Reduces risk of culling 95% CI = 1.2 to 3.6, P = 0.01 Assumptions: Partial Budget Fresh cows 100 SCK incidence (%) 40 BHBA strip cost ($) 1.3 Labor ($/hr) 10 Testing (cows/hr) 45 Treatment (cows/hr) 60 Propylene Glycol ($/dose) 1 Number days of treatment 5 DA (%) 5 Cost of a DA 300 Early cull (%) 10 Replacement 750 Price of milk ($/100 lb) 20 Feed ($/lb) 0.11 Partial Budget / 100 fresh cows No Testing Expenses: Revenues: Labor test 0 Milk 720 Labor tx 83 DA 563 PG 500 Removal 1571 Strips 0 Feed 198 Outcome: 2073 Testing until positive with 5 days PG Expenses: Revenues: Labor test 112 Milk 720 Labor tx 33 DA 563 PG 200 Removal 1571 Strips 103 Feed 198 Outcome: 2208 Cows calve a year 100 SCK incidence (%) 40 BHBA strip cost ($) 1.3 Labor ($/hr) 10 Testing (cows/hr) 45 Treatment (cows/hr) 60 PG ($/dose) 1 Number days of tx 5 DA (%) 5 Cost of a DA 300 Early cull (%) 10 Replacement 750 Price of milk ($/100 lb) 20 Feed ($/lb) 0.11 Partial Budget / 100 fresh cows SCK Incidence Acknowledgements 20% 40% 60% No Testing $ 1,026 $ 2,073 $ 3,119 Cornell vet students Wisconsin vet students Study herds Test until positive $ 1,307 $ 2,208 $ 3,108 This project was supported by Agriculture and Food Research Initiative Competitive Grant no from the USDA National Institute of Food and Agriculture. 18