Effect of Prepartum Dry Matter Intake on Liver Triglyceride Concentration and Early Lactation

Effect of Prepartum Dry Matter Intake on Liver Triglyceride Concentration and Early Lactation SANDRA J. BEFITICS, RIC R. GRUMMER,' CARLOS CADORNIGA-VALINO, and EMILY E. STODDARD Dairy Science Department University of Wisconsin Madison 53706 ABSTRACT

INTRODUCTION

Depression in feed intake during the final week before calving was hypothesized to be a major factor in the etiology of fatty liver development near parturition. Eleven cows were allowed to eat for ad libitum intake prior to calving (control), and 11 cows were maintained at the same level of DMI recorded during d 21 to 17 prior to calving by force feeding the feed refusals via rumen cannulas. Feed intake by control cows decreased 28% during the final 17 d prior to calving. Lipid triglyceride increased 227 and 75% for control and force-fed cows between d 17 prior to parturition and d 1 following calving. Dry matter intake prior to calving was correlated negatively with liver triglyceride immediately after calving (r = -30). Plasma glucose concentrations for control and force-fed cows were 63 and 76 mg/dl2 d prior to calving and also were related closely to liver triglyceride immediately after calving (r = -SO). By d 28 after calving, there were no differences in liver triglyceride between treatments. Cows that were forcefed prior to calving tended to yield milk with greater fat percentage (4.22 vs. 3.88%) and to yield more 3.5% FCM (46.1 vs. 41.7 kg/ d) during the first 28 d postpartum. (Key words: prepartum, feed intake, liver triglyceride, dairy cows)

Fatty liver is a metabolic disorder that occurs when the rate of fatty acid uptake and esterification exceeds the rate of fatty acid depletion either through oxidation or export as triglyceride (TG) within very low density lipoproteins. Fatty liver is perceived by many to develop during the period immediately postpartum when the cow is experiencing negative energy balance ftom inadequate feed intake. Negative energy balance results in fatty acid mobilization from adipose tissues and elevated plasma NEFA. Hepatic fatty acid uptake is related to plasma NEFA concentration (4). Reid (21) indicated that 66% of Friesian cows had moderate to severe fatty liver at wk 1 after calving. Several studies indicate that fatty liver may develop prepartum. Gerloff et al. (9) demonstrated that cows prone to severe fatty liver exhibited elevated hepatic TG prior to calving. More recently, we observed that liver TG increased from 2 to 17% (DM basis) between d 17 prior to calving and d l to 2 postpartum in Holstein cows (24). At 4 wk postpartum, liver TG was slightly less than immediately after calving. Liver TG may peak between 0 and 4 wk postpartum (11); however, significant lipid infiltration occurs by calving. Ruminant animals have an inherently low rate of very low density lipoprotein secretion relative to many species of animals (16, 20). Undoubtedly, this is a major factor contributing to the development of fatty liver. Little research has been conducted to identify other Abbreviation key: BHBA = fLhydroxybutyr- factors that potentiate the development of ate, C = control, FF = force-fed, TG = trigly- periparturient fatty liver. Changes in endocrine status of the animal as parturition approaches ceride. may be important. Estrogen, a potent regulator of hepatic fatty acid metabolism in nonruminants, increases in plasma prepartum and may Received January 6, 1992. play a role in fatty liver development (10). Dry Accepted March 23, 1992. 'Reprint requests: 1675 Observatory Drive, Madison, matter intake decreases in dairy cattle immediately prepartum, particularly during the final 5 WI. 1992 J Dairy Sci 75:191&1922

1914

PREPARTUM FEED INTAKE E!FFE€T ON FA"Y LIVER

to 7 d prior to calving [(5); Skaar et al., 1989, University of Wisconsin-Madison, unpublished data]. The depression in DMI may be related to endocrine changes (10) and may account for increased plasma NEFA concentration prepartum (Skaar et al., 1989, University of Wisconsin-Madison, unpublished data). An increase in plasma NEFA and hepatic fatty acid uptake prepartum could result in development of fatty liver if there is not a commensurate increase in fatty acid oxidation, very low density lipoprotein TG secretion, or both. The objective of this experiment was to determine the contribution of feed intake depression prior to calving on the development of fatty liver. A second objective was to determine the effect of prepartum feed intake on lactation performance immediately following parturition. MATERIALS AND METHODS Experimental Deslgn

and Sampling Protocol

Pregnant, nonlactating cows (n = 22) that had completed at least two lactations were paired according to calving date and 305-d mature equivalent milk yield that was computed following the previous lactation, AU cows began the experiment at d 31 prior to expected calving, were fed forage for ad libitum intake for 14 d, and received vitamins and minerals according to NRC (19) recommendations using wheat middlings as carrier. Forage consisted of 50% alfalfa silage and 50% corn silage (DM basis). During d 21 to 17 before calving, average daily DMI was determined for all cows. Beginning d 16 prior to calving until d 1 following calving, all cows were offered the average amount of DM that they consumed during d 21 to 17 prior to calving. One cow from each pair was force-fed (FF)until 1 or 2 d after calving by placing daily feed refusals into the rumen through a cannula after wetting with warm water. The other cow from each pair was not cannulated and was allowed to experience the feed intake depression that often occurs near calving (control treatment; C). Cows were offered a diet consisting of 25% concentrate and 75% forage (DM basis) for ad libitum intake from d 1 or 2 through d 5 postpartum and 50% concentrate and 50% for-

1915

age (DM basis) during d 6 through 28 postparh1.m. Concentrate included the following: 43.40% course ground corn; 16.00%corn gluten meal; 14.00% corn gluten feed, 10.00% soybean meal; 10.00% wheat middlings; 4.00%meat meal; .60% vitamins A, D, and E (2,643,171 lU of vitamin A, 660,792 IU of vitamin D, and 661 IU of vitamin E); 1.00% limestone; and 1.00% trace-mineralized salt. Forage consisted of 50% alfalfa silage and 50% corn silage (DM basis). Crude protein and NDF percentages of the diets were estimated from near infrared analysis of forages (University of Wisconsin Plant and Soils Analysis Laboratoxy, Madison) and NRC (19) values for concentrate ingredients and were 13.9, 41.5, 17.3, 35.5, 20.8, and 29.5 when forage to concentrate ratios were 100:0, 7525, and 5 0 50. Net energy for lactation for the diets was based on NRC (19) values for all feeds, and means were 1.51, 1.60, and 1.69 Mcal/kg for the respective forage to concentrate ratios. Blood from the coccygeal vein was obtained 17 d prior to expected calving, once during the week prior to calving, and on d 1 or 2, 14, and 28 following calving. Liver biopsies (24) were obtained 17 d prior to calving and on d 1 or 2 and 28 following calving. On days that blood and liver samples were obtained, blood was drawn immediately prior to the liver biopsy. Most but not all of the samples taken immediately after calving were obtained within 24 h after parturition. For cows receiving C, 9 of the blood and liver samples were obtained 1 d after calving, and 2 were obtained 2 d after calving. For FF cows, 8 of the cows were sampled 1 d after calving, and 3 were sampled 2 d after calving. Force feeding was continued until the first biopsy postpartum; all cows were allowed ad libitum consumption of diets thereafter. Plasma was obtained from blood and analyzed for P-hydroxybutyrate (BHBA), NEFA, and glucose according to Skaar et al. (24). Liver tissue was analyzed for total lipid (24), TG (24). and glycogen (15). Milk yield was recorded daily. Milk samples were taken at a.m. and p.m. millcings on d 6,7,13,14,20,21,27, and 28 postpartum and analyzed for percentages of fat and protein and for somatic cells by infrared spectrophotometry (Wk.consin DHI Cooperative Laboratory, Appleton). Body condition was evaluated by two people on d 17 prior to expected calving. Journal of Dairy Science Vol. 75, No. 7, 1992

1916

BERTICS ET AL.

Body weights were recorded on d 18, 17, 11, 10,4, and 3 prior to calving and on d 6.7, 13, 14,20,21,27, and 28 following Calving. Samples of corn and alfalfa silages were obtained every 2nd wk throughout the trial, when new silos were started, and when changes in silage DM were noted. Dry matter content of forages and concentrates was determined by ovendrying at W C . Statistical Analysis

Data for liver and plasma metabolites were analyzed within sample time according to the following model:

where Yi = the dependent variable, p = the overall mean of the population, b = the regression coefficient for the liver or plasma metabolite measured at d 17 prior to expected calving for cow i, Ti = the average effect of treatment i, and ei = the unexplained residual element assumed to be independent and identically distributed.

Data for BW, DMI, and milk yield and composition were analyzed according to the following model:

(i.e., block) originally was included in all statistical models; however, the term was eliminated because it did not increase the sensitivity of the models to determine a significant treatment effect. Variation between cows was used as the error term when testing for treatment effects on BW, DMI, and milk yield and composition. Treatment differences or treatment by time interactions were considered to be significant at P < .05, to have a tendency toward significance at P = .05 to P = .15, and to be nonsignificant at P > .15. One FF cow was removed during wk 1 postpartum because of severe mastitis. Data from a liver biopsy obtained 1 d postpartum from an FF cow were removed from the statistical analysis because they were outliers. Mean, range, and standard deviation for liver total lipid percentage (DM basis) for 10 FF cows (excluding the outlier) at 1 d postpartum was 21.4, 14.4 to 27.8, and 3.7. When all 11 cows (including the outlier) receiving this treatment were considered, these values were 25.6, 14.4 to 67.1, and 14.2. Because the liver sample that contained 67.1% total lipid was about 12 SD away from the mean calculated for FF cows without the sample, it was considered an outlier. The cow from which the liver sample containing 67.1% lipid was taken delivered twins. However, we cannot conclude that twinning per se was the causative factor for the unusually high value, because a C cow had twins and had only 27.1% total lipid in the first liver sample taken after calving. RESULTS AND DISCUSSION

where

Pretreatment Through d 1 Postcalvlng

A summary of previous lactation data indicated that the two treatment groups were relatively balanced for milk yield and for fat and protein percentages (Table 1). Body condition of cows from each treatment was similar at d 17 prior to expected calving. Body weights and BW changes (Figure 1) during the trial were not significantly different between treatments. Cows from each group were on treatment for Period represented daily measurements for approximately 20 d prior to calving, but there DMI and milk yield and weekly measurements was considerable variation among cows within for BW, milk composition, and FCM. Prepar- each treatment. The final prepartum blood tum and postpartum data were analyzed sepa- sample was obtained, on the average, 2.1 and rately for BW and DMI measurements. Pair 1.3 d prior to calving for C and FF cows. The

= the dependent variable, = the overall mean of the population, = the average effect of treatment i, = the average effect of cow j, Pk = the average effect period k, and q$ = the unexplained residual element assumed to be independent and identically distributed.

Yi$ p Ti Cj

Journal of Dairy Science Vol. 75, No. 7, 1992

1917

PREPARTUM FEED INTAKE FFFJXT ON FATTY LIVER

TABLE 1. Characteristics of

cows used in the study and prepartum sampling times.

a

b

Initiation of treatment, d prepartum 19.9 BW @retreatment),kg 730 3.59 Body condition' (preeeabnent) Final prepartum blood sample, d PEP2.1 Previous lactation data3 305-d ME Milk, kg 9586 Fat, % 351 Protein, % 3.12

SD

e

5-33 643-7% 3.25-4.00

7.9 53 25

1.&5.0

1A

8181-11.014 3.07-3.81 2.893.47

871 22 .ia

-

Range

SD

9-47 697-881 3.00-3.80

9.7 56 .26

1.o-3.0

1 .o

X

20.2 794 3.61

1.3 9758 3.62 3.13

8984-11.409 3.18404 2.92-329

776

.24 .14

'C = Conlrol cows were allowed to experience feed iutake depression prior to calving, FF = cows were force-fed orts via rumen cannula prior to calving. 'Scale ranged from 1 = thin to 5 = obese. 'ME = Matare equivalent.

difference between treatment groups was greater than desired during t h i s time because significant endocrine changes associated with parturition could confound the data. However, given the difference, it is preferable to sample FF cows closer to calving. This probably results in a bias toward greater plasma NEFA concentrations in the samples obtained immediately prepartum in FF cows, because plasma NEFA concentrations increase as parturition approaches. Dry matter intake was significantly lower prepartum for C cows, and treatment by time interaction prior to calving (Figure 2) was significant. Feed intake of C cows began to

2

80

-4

-2

0

2

4

Week Relative to Calving Figure 1. Body weight changes of cows allowed to experience feed intake depression prior to calving (0)or those force-fed (9orts via rumen cannula prior to calving.

decline most dramatically beginning 1 wk prepartum and was reduced by approximately 30% from pretreatment feed intake. This was similar to the depression in prepartum feed intake observed in a previous trial (Skaar et al., 1989, University of Wisconsin, unpublished data) and similar to that reported by others (5, 14,17). Coppock et al. (5) indicated that prepartum DMI depression may be greater for cows consuming high grain diets than for those consuming high forage diets. We noted that FF cows also experienced a depression in feed intake for the first 2 d after calving when orts were no longer placed in the rumen. This suggests that factors other than gut fii contrib-

- 2 5 - 2 0 -15 -10 - 5

0

5

10 1 5 2 0 25 3 0

Day Relative to Calving Figure 2. Dry matter intake of cows allowed to experience feed intake depression prior to calving (0) or those force-fed (0) orts via rumen cannula prior to calving. Journal of Dairy Science Vol. 75, No. 7, 1992

1918

BERTICS ET AL.

TABLE 2. Effect of prepartum DMI on liver total lipid, triglyceride, and glycogen concentrations (DM basis).' Liver lipid, % Relative to parturition d -17 (co~ariateperiod) d l d 28 Liver triglyceride, % d -17 (covariate period) d l d 28 Liver glycogen,3 % d -17 (covariate period) d l d 28

C*

SE

ppz

SE

P

15.9 30.7 30.6

.6 2.6 5.7

23.5 35.1

2.7 5.7

.a8 .60

7.1 23.2 26.9

3.0 6.3

12.4 25.3

3.2 6.3

.03 .85

4.2 2.7

1.2 .7

.33 .41

1.3 2.5 3.6

.8

.I 1.7 .8

'Data for d -17 are pooled because cows were not on treatment, and data for d 1 and 28 are covaxiately adjusted means. 2C = Control cows were allowed to experience feed intake depressionprior to calving; FF = cows were force-fed orts via m e n cannula prior to calving. 31nsufficient tissue was available to perfom analysis on all samples. n = 9, 11, and 9 for C and 10, 9, and 9 for FF cows on d -17, 1, and 28, respectively.

Ute to regulation of feed intake near the time of calving, because FF cows should have had sufficient space in the rumen to maintain feed intake at prepartum levels. Liver total lipid and TG contents were 15.9 and 7.1% (DM basis) prior to initiation of treatments (Table 2). Total lipid content was similar, and TG content was higher, than previously measured at a similar time prepartum (24). Immediately following calving, liver total lipid tended to be higher in cows that were allowed to experience a reduction in feed intake, and liver TG was also significantly higher in these cows. Liver TG would be expected to be a more sensitive measure of fatty liver than liver total lipid because TG synthesis and storage are the predominant metabolic fate of fatty acids when hepatic capacity for oxidation is exceeded. Our data contrast with those of Drackley et al. (6). who indicated that feed restriction to approximately 80% of ad libitum consumption did not result in fatty liver in early lactation cows, even though plasma NEFA concentrations were greatly increased. However, in their study (6), liver TG probably already was elevated when feed restriction was introduced at 14 d postpartum. Liver TG increased 227% for C cows versus 75% for those kept on feed during the final 3 wk prepartum. The increase in liver TG Journal of Dairy Science Vol. 75. No. 7, 1992

observed for FF cows indicates that additional factors beyond feed intake also may contribute to the development of fatty liver. Endocrine changes at calving may be involved, and we have previously suggested a role for estrogen, which increases near calving (10). Low Serum insulin and elevated lipolytic hormones, such as somatotropin, placental lactogen, and prolactin, near calving may be involved (12). Some of the orts that were placed into the rumen of FF cows probably accumulated in the rumen. In 2 of the FF cows, we were not able to fit all of the orts into the rumen during the last 2 to 3 d prior to calving because of limited space. Also, a possible depression in digestibility of feed passing out of the rumen of FF cows makes it possible that digestible energy intake was depressed even though DMI was maintained. A depression in digestible energy intake by FF cows near parturition may have contributed to the elevation in liver TG at calving. Plasma NEFA concentration increased between d 17 prepartum and calving, but there were no significant differences between treatments at -2 or +1 d relative to calving (Table 3). Plasma NEFA concentration averaged 235 higher in C cows 2 d prior to calving, but variation among cows was too great for us to detect a difference. Plasma NEFA concentra-

1919

PREPARTUM FEED INTAKE EFFECT ON F A T N LIVER

TABLE 3. Effect of prepartum DMI on plasma NEFA, bhydroxybutyrate (BHBA), and glucose concentrations.' C2

SE

FF2

SE

P

385 876 992 500 395

51 123 105 74 68

641 1064 824 534

129 111 82 76

20 .64 .o 1 .19

~

NE% PJU Relative to parturition d -17 (covariate perid) d -2 d l d 14 d 28 Glucose, mddl d -17 (covariate period) d -2 d l d 14 d28 BHBA, mg/dl d -17 (covariate perid) d -2 d l d 14 d 28

70.3 63.4 60.3 55.6 56.7

2.4 3.3 3.5 3.4 3.0

76.5 59.0 50.1 59.2

3.5 3.6 35 3.2

.80 .28 .5 8

8.O 11.9 17.6 17.1 14.7

.9 1.3 2.6 5.7 3.6

12.5 18.1 32.4 18.2

.13 2.8 6.0 3.4

.77 .89 .10 .53

.a?

'Data for d -17 are pooled because cows were not on treatment, and data for d 1 and 28 are covariately adjusted means. *C= Control cows were allowed to experience feed intake depression prior to c a l w , FF = cows were force-fed orts via m e n cannula prior to calving.

tion increased rapidly during the final days preceding parturition (24); therefore, the large variation in NEFA probably was a result of variation in the time of the final blood sampling prior to calving. A dramatic difference between plasma glucose concentrations in C and FF cows at d 2 prior to parturition was observed. Glucose concentrations usually do not vary greatly because of nutritional modifications. The increase in plasma glucose for FF cows was particularly impressive, considering that abomasal glucose or propionate infusions may not alter plasma glucose concentration as dramatically (7). Rates of hepatic gluconeogenesis decrease in response to exogenous glucose administration to lactating cows (3). The reason or reasons for the increase in plasma glucose for FF cows are unhown but presumably are related to the higher nutrient intake. There were no differences in blood glucose between treatments the day after calving. Plasma BHBA concentrations were not different between treatments. Elevated ketone concentrations in blood usually coincide with elevated blood NEFA and low blood glucose concentration. Hepatic ketogenesis probably

increases because of an enhanced rate of fatty acid oxidation and gluconeogenesis. The later may act to deplete tricarboxylic acid cycle intermediates, resulting in incomplete oxidation of fatty acids and ketone synthesis (2). Plasma NEFA and BHBA increased as calving approached and were higher on the day after calving than at 2 d prior to calving. However, by the day after calving, there were no differences in plasma glucose between treatments, which may partially explain the absence of a difference in BHBA concentrations between treatments at that time. Veenhuizen et al. (25) indicated that cows clinically induced to develop ketosis with moderate feed restriction and with feeding of 1.3-butanediol exhibited elevated liver TG and reduced liver glycogen prior to the development of clinical ketosis. In light of reduced liver TG and elevated prepartum glucose concentrations in FF cows, we decided to analyze liver glycogen. hitially, we did not plan to measure liver glycogen; therefore, for some cows, liver tissue was insufficient to perform the glycogen assay. Consequently, n = 9, 11, and 9 for C cows and 10,9, and 9 for FF cows on d -17, +1, and +28 relative to calving, Journal of Dairy Science Vol. 75, No. 7, 1992

1920

BERTICS ET AL.

respectively (Table 2). Average liver glycogen was higher for FF cows the day after calving, but the difference was not significant. Correlation analysis was performed to identify parameters that might be related to liver total lipid and TG concentrations. Dry matter intake was most highly related to liver TG and total lipid (Table 4). There was substantial variation within the C group in the extent of feed intake depression prior to calving; therefore, we computed a correlation coefficient between DMI and liver TG within this treatment group alone. The correlation also was highly negative and significant when only those cows were considered (r = -.79, P < .01). Plasma glucose at d 2 prior to calving was correlated negatively with liver TG 1 d postpartum. Although treatment differences were not observed for NEFA and BHBA, plasma NEFA at -2 d tended to be related positively to liver total lipid measured the day after calving, and NEFA and liver TG tended to be positively related at 1 d after calving. A positive correlation between plasma NEFA and liver total lipid following calving previously was reported (22). Plasma BHBA at -2 d tended to be positively related to liver TG and total lipid immediately after calving.

Postpartum Performance

At 28 d after calving, there were no significant differences in liver total lipid, TG, or glycogen (Table 2). Plasma NEFA were significantly higher in FF cows on d 14 postpartum (Table 3). Coupled with a lower average plasma glucose in FF cows,this may account for the trend toward higher BHBA concentrations in plasma on d 14 postpartum. Higher energy expenditures for milk yield (Figure 3 and Table 5 ) by FF cows may account for the higher plasma NEFA and BHBA on d 14 postpartum. There was no significant main effect of prepartum DMI on milk yield, but there was a trend for a time by prepartum treatment interaction and a main effect of prepartum treatment on milk fat percentage and 3.5% FCM yield. It cannot be determined from this study whether liver TG content at calving directly affects lactation performance. Little is known about the effects of cellular TG accumulation on hepatocyte function. Liver slices from cows with elevated liver TG from moderate feed restriction and 1,3-butanediol administration had a lower rate of glucose synthesis from propionate (18). Similarly, hepatocytes in suspension obtained from goats in negative energy balance and receiving only

TABLE 4.Correlation coefticients between liver total lipid or triglyceride (TG) percentages at d 1 and 28 postpartum and DMI, FCM yield, plasma metabolites, and liver glycogen. postpartum

Parameter DMI

NEFA NEFA Glucose Glucose

BHBA* BHBA

DMI 3.5% FCM 3.5% FCM/DMI

NEPA Glucose

BHBA Liver glycogen

Time relative to calving (4 -1 -2 1 -2 1 -2 1

28 28 28 28 28 28 28

'BHBA = FHydroxybutyrate. Journal of Dairy Science Vol. 75, No. 7, 19!32

Liver TG

r

Liver lipid

r

P

.02

-.69 .41 .34 -28

.80

.31

.ooo4 .06 .13 21 .17

.06 51

.37 26

26

P

d l -.SO

.OOO1

28 .40 -50 -.06 .41 .I5

22

-.66

.002 .I2

-.35

56 .31 -.41 .61 -.72

.w

.01 .19

-64 -.33 52 28

.07

-.a

.005

.63 -.71

.001

.09

.003 .16 .02 24 .05

.ow .001

PREPARTUM FEED

1921

INTAKE EFFJ3X ON FATIY L.IVER

TABLE 5. Postpartum DMI and lactation performance.

DhlI, kg/d Millr;' W d 35% FCM, kg/d

Fat % Pmteh %

Fat, kg/d

Protein. W d SCC, 1 O o o m 20

I

0

5

10

15

20

25

30

Day Relative to Calving Figure 3. Milk yield of cows allowed to experience feed intake depression prior to calving (0)or those force fed orts (0) via rumen cannula prior to calving.There was a trend toward a treatment by time interaction (P < .12).

C'

FF

P

20.5 36.7 41.7 3.88 3.16 1.51 1.24 553

19.9 39.1 46.1 4.22 3.23 1.73 1.33 338

.62 .30 .ll .06

.35 .02 .33 .46

IC = Control cows were allowed to experience feed intake depression prior to tal-, FF = cows were forcefed orts via rumen cannula prior to calving. %rmd toward significant prepartum treatment by time interaction (P < .12).

ably prior to calving. Overconditioned cows have lower appetite following calving (8, 13). protein via stomach tube had a lower rate of We do not know whether body condition influglucose synthesis from propionate compared ences prepartum appetite. Our data did not with cells from full-fed goats (1). One might reveal a significant correlation between body speculate that hepatic glucose synthesis may condition and prepartum intake; however, this be limiting for milk yield in early lactation, may have been a consequence of the limited particularly if curtailed because of excessive number of cows on experiment and the relahepatic TG storage. However, in the current tively uniform body condition of the cows. study, differences in milk yield between treat- Enhancing the carbohydrate status of the cow, ments began to appear when treatment differ- in particular by the addition of glucose, during the final week prior to freshening also may be ences in liver TG were disappearing. beneficial. hcreasing n o n s t r u c w carbohyAt 28 d postpartum, liver glycogen was the drate intake by grain feeding adjustment may parameter that was most highly related to liver have merit. However, limitations to the amount total lipid and TG content (Table 4). This confirms the relationship between these two of grain that should be fed prepartum and the parameters that was documented previously inherent decline in feed intake that occurs prior (25) during a ketosis induction protocol. The to calving may restrict the benefit that could be correlation between liver TGglycogen ratio obtained via this feeding strategy. Administraand BHBA was .58 (P c .Cn), which also tion of glucose precursors during the final days supports the contention that the ratio is an prior to calving may be advantageous. Increasing blood glucose may elicit an insulin reimportant factor in the etiology of ketosis (25). sponse and reduce fatty acid mobilization from Dry matter intake at 28 d postpartum also was adipose. Propylene glycol administration to correlated highly with liver total lipid and TG, lactating cows successfully increased blood even more so than gross efficiency of milk glucose and lowered blood ketones (23). We yield (3.5% FCMDMI). Plasma glucose was curzently are examining the effects of propyrelated negatively and BHBA was related posi- lene glycol administration during the final 5 d tively to liver total lipid and TG. prior to calving on the development of fatty liver, plasma ketones, and postpartum lactation performance. CONCLUSIONS

Data from this experiment suggest that DMI is a critical factor influencing the development of fatty liver. Further research should investigate means to maximize feed intake immedi-

ACKNOWLEDGMENTS

Leland Danz supervised the care, feeding, and force-feeding of cows that was vital to the Jonmal of Dairy Science Vol. 75, No. 7, 1992

1922

BERTICS ET AL.

completion of this experiment. His efforts were deeply appreciated by the authors. REFERENCES 1h e n t a n o , L. E., R R. Onrmmer, S. J. Bertics, T. C. Skaar, and S. S. DoW. 1991. Effects of energy balance on hepatic capecity for oleate metabolism and triglyceride secretion. J. Dairy Sci. 74132. 2Baird, G. D., R. J. Heitzman, I. M. Reid, H. W. Symonds, and M. A. Lomax. 1979. Effects of food deprivation on ketonemia, ketogenesis and hepatic intermediary metabolism in the nonlactating cow. Biochem. J. 17835. 3Bartley. J. C., and A. L. Black. 1966. Effect of exogenous glucose on glucose metabolism in dairy cows. J. Dairy Sci. 60:204. 4 Bell, A. W. 1980. Lipid metabolism in the liver and selected tissues and in the whole body of ruminaat animal. Prog. Lipid Res. 18:117. SCoppock, C. E., C. H. Noller, S. A. Wolfe, C. J. callhan, and J. S. Baker. 1972. Effect of forage: concentrate ratio in complete feeds fed ad libitum on feed intake prepartum and the occumnce of abomasal displacement in dairy cows. J. Dairy Sci. 55783. 6 Drackley, J. K., J. J. Veenhuizen, M. J. Richard, and J. W. Young. 1991. Metabolic chauges in blood and liver of dairy cows during either feed restriction or .. adrrmustration of 1,3-butanediol. J. Dairy Sci. 7 4 4254. 7Frobish, R A., and C. L. Davis. 1977. Effects of abomasal infusions of glucose and propionate on milk yield and composition. J. Dairy Sci. 60:204. 8 Gamsworthy. P. C., and J. H. Topps. 1982. The effect of body condition of dairy cows at calving on their food intake and performance when given complete diets. Anim. Prod. 35:113. 9Gerloff, B. I., T. H. Herdt. and R S. Emery. 1986. The relationship of hepatic lipidosis to health and performame in dairy cattle. J. Am. Vet. Med. Assoc. 188:845. lOGrummer, R. R., S. J. Bertics, D. W. Lacount, J. A. Snow, M. R. Denthe, and R H. Stauffacher. 1990. Estrogen induction of fatty liver in dairy cattle. J. Dairy Sci. 73:1537. IlHanison, R. 0.. S. P. Ford, J. W. Young, A. J. Conley, and A. E. Freeman, 1990. Increased milk production versus reproductive and energy status of high producing dairy cows. I. Dairy Sci. 73:2749. 12 Herdt, T. H. 1988. Fatty liver in dairy cows.Vet. Clin. North Am. Food Anim. Ract. 4269.

Journal of Dairy Science Vol. 75, No. 7, 1992

13Holter, J. B., M. J. Slotnick. H. H. Hayes, C. K. Bo& W. E. Urban,Jr., and M L. McGilliard. 1990. Effect of prepartum dietary energy on condition score, postpartom energy, nitrogen partitions, and lactation production responses. J. Dairy Sci. 73:3502. 14Journet, M., and B. Remod. 1976. Physiological factors affecting the voluntary intake of feed by cows: a review. Livest. Prod. Sci. 3129. 15 Keppler. D., and K. Decker. 1983. Glycogen. Page 11 in Methods of Enzymatic Analysis. 3rd ed. Vol. VI. Metabolites 1: Carbohydrates, H. U. Bergmeyer. ed. Academic Press, New York, NY. 16 Kleppe, B. B., R. J. Aiello, R R. Grummer, and L. E. Armentano. 1988. Triglyceride accumulation and very low density lipoprotein secretion by the rat and goat hepatocytes in vitro. J. Dairy Sci. 71:1813. 17Lodge, G. A., L. J. Fisher, and J. R Lessard. 1975. Influence of prepartum feed intake on performance of cows fed ad libitum during lactation. J. Dairy Sci. 58: 696. 18Mills, S. E., D. C. Beik, and J. W. Young. 1986. Evidence for impaired metabolism in liver during induced lactation ketosis of dairy cows. J. Dairy Sci. 69:362. 19National Research Council. 1989. Nutrient Requirements of Dairy Cattle. 6th d.Natl. Acad. Sci., Washington$ Dc. 20 Pullen, D. L., J. S. Liesman, and R. S. Emery. 1990. A species comparison of liver slice synthesis and sec~etion of triacylglycerol from nonesterifkd fatty acids in media. J. Anim. Sci. 68:1395. 21 Reid, I. M. 1980. Incidence and severity of fatty liver in dairy cows. Vet. Rec. 107:281. 22Reid, I. M., G. L. Rowlands, R A. Collins. C. J. Roberts, and R Manston. 1983. The relationship b s tween post-parturient fatty liver and blood composition in dairy cows. J. Agric. Sci. (Camb.) 101:473. 23Sauer, F. D., J. D. Erfle, and L. J. Fisher. 1973. Propylene glycol and glycerol as a feed additive for lactating dairy cows: an evaluation of blood metabolite parameters. Can. J. Anim. Sci. 53:265. 24 Skaar, T. C., R R. Grummer, M. R Dentine, and R H. Stauffacher. 1989. Seasonal effects of prepartum and postpartum fat and niacin feeding on lactation performance and lipid metabolism. J. Dairy Sci. 7 2 2028. 25 Veenhuizen, J. J., J. K. Drackley, M. J. Richard, T. P. Sanderson, L. D. Miller. and J. W. Young. 1991. Metabolic changes in blood and liver during develop m a t and the treament of experimental fatty liver and ketosis in cows. J. Dairy Sci. 744238.