J. Dairy Sci. 89:2141-2157
© American Dairy Science Association, 2006.
Prepartal Plane of Nutrition, Regardless of Dietary Energy Source, Affects Periparturient Metabolism and Dry Matter Intake in Holstein Cows1
G. N. Douglas2,
T. R. Overton3,
H. G. Bateman, II4,
H. M. Dann5 and
J. K. Drackley6
Department of Animal Sciences, University of Illinois, Urbana 61801
6 Corresponding author: drackley{at}uiuc.edu
 |
ABSTRACT
|
|---|
Previous research in our laboratory showed that dietary fat supplementation during the dry period was associated with decreased peripartum hepatic lipid accumulation. However, fat supplementation decreased dry matter (DM) intake and thereby confounded results. Consequently, 47 Holstein cows with body condition scores (BCS) 
3.5 at dry-off were used to determine whether source or amount of energy fed to dry cows was responsible for the decreased hepatic lipid content. Moderate grain- or fat-supplemented diets [1.50 Mcal of net energy for lactation (NEL)/kg] were fed from dry-off (60 d before expected parturition) to calving at either ad libitum (160% of NEL requirement) or restricted (80% of NEL requirement) intakes. Postpartum, cows were fed a single lactation diet for ad libitum intake and performance was measured for 105 d. Prepartum intakes of DM and NEL were significantly lower for feed-restricted cows as designed. During the first 21 d postpartum, previously restricted cows had higher intakes of DM and NEL. Body weights and BCS were lower prepartum for restricted cows but groups converged to similar nadirs postpartum. Restricted-fed cows had lower concentrations of glucose and insulin and increased concentrations of NEFA in plasma during the dry period. Peripartum NEFA rose markedly for all treatments but were higher postpartum for cows previously fed ad libitum. Plasma concentrations of NEFA and BHBA remained lower in cows restricted-during the dry period. Postpartum concentrations of total lipid and triglyceride in liver were lower in cows previously feed-restricted. Across dietary treatments, activity of carnitine palmitoyltransferase (CPT) in hepatic mitochondria was lowest at – 21 d, highest at 1 d, and decreased at 21 and 65 d relative to parturition. The activity of CPT at d 1 tended to be higher for previously feed-restricted cows; thereafter, CPT activity declined more rapidly than in cows fed ad libitum. Nutrient intake during the dry period had more pronounced effects on peripartal lipid metabolism and DMI than did composition of the prepartum diet.
Key Words: dry period • feed restriction • liver metabolism • supplemental fat
|
INTRODUCTION
|
|---|
Optimal nutritional management during the dry period remains controversial. Successful programs would minimize health disorders and maximize productivity in the subsequent lactation. Because of the association between hepatic lipid accumulation and periparturient disorders (Grummer, 1993; Rukkwamsuk et al., 1998; Bobe et al., 2004), nutritional strategies should promote vigorous appetites after parturition that result in rapid increases of DMI, which in turn will modulate body lipid mobilization and result in lower accumulation of triglyceride (TG) in liver. Body condition of cows at parturition can impact peripartum DMI and subsequent performance (Fronk et al., 1980; Garnsworthy and Topps, 1982). Maximal reproductive performance and milk production may not be achieved if BCS is inadequate (Waltner et al., 1993). Consequently, dry period nutrition that allows some BCS gain has generally been considered beneficial. On the other hand, overfeeding that allows excessive BCS gain has been clearly shown to increase the incidence of peripartal health problems (Fronk et al., 1980; Grummer, 1993; Rukkwamsuk et al., 1998). However, most studies have concentrated on the effects of overconditioning on peripartum performance rather than on obtaining optimal conditioning in thinner cows.
In previous research from our laboratory, we attempted to add BCS during the dry period by increasing the energy density of the dry diet either by additional cereal-based concentrates or by an isocaloric addition of supplemental fat to a high forage diet (Grum et al., 1996). Cows fed the high forage diet supplemented with fat throughout the dry period exhibited markedly lower total lipid and TG concentrations at d 1 postpartum compared with cows fed the low-energy control or the high grain diets. Lower TG accumulation was associated with decreased capacity of liver slices to esterify palmitate and greater capacity for ß-oxidation of palmitate by liver homogenates. However, fat-fed cows consumed significantly less DM and lost BCS during the dry period compared with cows fed the control or high grain diets (Grum et al., 1996). Thus, whether the lower liver TG content was due to the fat supplementation or to the lower nutrient intake could not be determined. In a companion experiment to the present study (Douglas et al., 2004), we found that neither isocaloric addition of fat nor addition of fat to increase dietary NEL density during the dry period affected peripartal lipid accumulation in liver or subsequent milk production. Therefore, the primary objective of the present study was to determine if the metabolic changes resulting in the markedly lower accumulation of total lipid and TG at calving observed by Grum et al. (1996) were attributable to voluntary nutrient restriction resulting from decreased DMI or to a potential interaction of dietary fat and decreased DMI. Secondary objectives were 1) to determine if dry period feeding strategy affects postpartum DMI and lactation performance, and 2) to determine the effects of dry period diet and feeding strategy on other aspects of metabolism, including activity of carnitine palmitoyltransferase (CPT), which regulates entry of long-chain fatty acids (LCFA) into mitochondria for ß-oxidation and ketogenesis.
|
MATERIALS AND METHODS
|
|---|
Cows, Diets, and Experimental Design
All procedures were conducted under protocols approved by the University of Illinois Laboratory Animal Care Advisory Committee. Forty-eight Holstein cows were dried off 60 d before their expected parturition date and were assigned immediately and sequentially to 1 of 4 dietary treatment groups. Cows were chosen based on BCS 3.5 (1 to 5 scale) so that more conditioning might be added during the dry period. Diet formulation is shown in Table 1
. Diets fed during the dry period were based on alfalfa silage and corn silage, and consisted of a moderate-NFC control diet (C) and a low-NFC, fat-supplemented diet (F). The rumen-active fat source (Qual-Fat; National By-Products, Mason City, IL) consisted mainly of choice white grease. The LCFA composition as determined by GLC of methyl esters (Sukhija and Palmquist, 1988) was 3.1% C14:0, 0.9% C14:1, 24.2% C16:0, 4.9% C16:1, 0.3% C17:0, 13.8% C18:0, 46.2% cis-C18:1, 6.1% C18:2, and 0.4% C18:3. The fat had a calculated iodine value of 61 and an FFA content of 6.6%.
View this table:
[in this window]
[in a new window]
|
Table 1. Ingredient composition of control (C) and fat-supplemented (F) diets fed to dry cows and the lactation diet after parturition
|
|
The analyzed chemical composition of the diets is shown in Table 2. Diets C and F were formulated to provide adequate energy for a 650-kg cow with DMI of 12 kg/d (1.85% of BW) to gain about 36 kg of BW in excess of fetal growth, or about 0.6 BCS units (Otto et al., 1991), during the dry period. Diet F was formulated to be isocaloric to diet C by substituting oat hulls and fat for appropriate amounts of grain. Diets C and F were fed for either ad libitum (A) or restricted (R) intakes so that cows would consume approximately 160% of NEL requirements (A) or be limited to only 80% of NEL requirements (R). Therefore, dry period dietary treatments in a 2 x 2 factorial arrangement consisted of a moderate-NFC, ad libitumfed control group (CA), a moderate-NFC, restricted-fed group (CR), a fat-supplemented, ad libitumfed group (FA), and a fat-supplemented, restricted-fed group (FR). Groups CA and FA were the same as those used as part of the companion experiment reported previously (Douglas et al., 2004).
View this table:
[in this window]
[in a new window]
|
Table 2. Mean chemical composition and standard deviations of control (C) and fat-supplemented (F) diets fed during the dry period and the lactation diet fed after parturition
|
|
No transition diet was fed pre- or postpartum, but an anionic salt mixture based on fermentation by-products was added to all dry period diets at 9.1% of DM for 2 wk before the expected parturition date. Cows on feed-restriction treatments continued to be fed this reformulated close-up diet at the same restricted DMI through the day of calving. After parturition, all cows received a lactation diet (Table 1
) that contained 2.0% supplemental fat, and postpartum performance was measured for 105 d.
The number of cows per treatment was determined through power analysis (Morris, 1999) for the primary response variables, which were concentrations of total lipid and TG in liver tissue. In the study of Grum et al. (1996), mean concentrations of TG in liver (wet weight basis) at d 1 postpartum were 1.4, 7.3, and 5.9% for cows fed the fat-supplemented diet, control diet, and high-grain diet, respectively. Thus, the smallest difference between the fat-supplemented group and other treatments (4.5 percentage units) was > 90% of the overall mean TG content. Assuming similar variances as measured in that study, a difference of 3 percentage units (~60% difference) in hepatic TG concentration could be detected at P < 0.05 with 9 cows per treatment. Use of 12 cows per treatment in the present study had a predicted power to detect a difference in hepatic TG among groups (i.e., the interaction of intake and diet) of 50% with P < 0.05.
Cows were housed in tie stalls during the dry period until 2 wk before expected parturition, when they were moved to maternity box stalls until parturition. After calving, cows were returned to tie stalls. All cows were allowed to exercise daily in an outside lot from 0600 to 0930 h. Cows were milked twice daily and milk weights were recorded. Dry matter intake was measured daily and BW and BCS were measured weekly both prepartum and postpartum. The BCS were assigned by the same two individuals for the entire experiment, using a 1 = thin to 5 = obese scale (Wildman et al., 1982) with quarter-point increments. Health records were maintained for all cows and calf birth weights were recorded.
Sampling and Analysis of Feed and Milk
The TMR components were sampled weekly and dried at 55° C for determination of DM so that TMR composition and amounts of dietary DM offered could be adjusted if needed. Samples of TMR components and complete TMR were obtained weekly and composited monthly. Composites were analyzed using wet chemistry methods for contents of DM, CP, ADF, NDF, and minerals (Ca, P, Mg, and K; Table 2) by Dairy One Laboratory (Ithaca, NY). Lipid content was measured by diethyl ether extraction according to AOAC (1995).
Milk was sampled once weekly from consecutive a.m. and p.m. milkings. Samples were composited in proportion to milk production at each sampling and were preserved with 2-bromo-2-nitropropane-1,3-diol. Composited samples were analyzed for contents of fat and CP by infrared analysis (AOAC, 1995) at a commercial laboratory (Dairy Lab Services, Dubuque, IA).
Sampling and Analysis of Blood
Blood was sampled from the coccygeal vein or artery beginning at 0900 h; all sampling was completed before the morning dispensation of TMR. During the dry period, blood was sampled once weekly for 30 d after dry-off; twice weekly during wk – 4 and – 3, and 3 times weekly during wk – 2 and – 1 relative to expected parturition. After calving, blood was sampled 3 times weekly during wk 1 and 2 and twice weekly during wk 3 and 4; thereafter, blood was sampled once weekly until 105 d postpartum. Blood was not sampled on the day of calving. Blood samples were obtained before liver biopsy procedures if both occurred on the same day.
Blood samples were collected in evacuated test tubes (Vacutainer, Becton Dickinson Vacutainer Systems USA, Rutherford, NJ) containing sodium heparin. Plasma was obtained via centrifugation and aliquots were frozen at – 20° C for later analysis of the concentrations of glucose, insulin, NEFA, BHBA, total cholesterol, total protein, and urea N as described by Douglas et al. (2004).
Sampling and Analysis of Liver
Puncture biopsy was performed under local anesthesia as described in Douglas et al. (2004) to obtain approximately 3 to 5 g of liver tissue at – 65 and – 21 d relative to expected parturition and at 1, 21, and 65 d after parturition. Aliquots of liver tissue were immediately frozen in liquid N2 until later analysis for contents of total lipid, TG, and glycogen using methodology described in Douglas et al. (2004).
Another portion of the liver was homogenized with a motor-driven Potter-Elvehjem tissue homogenizer in a buffer containing 0.25 M sucrose, 20 mM Tris HCl (pH 7.4), and 1 mM Na-EDTA (van den Top et al., 1995). The mitochondrial fraction was isolated by differential centrifugation (van den Top et al., 1995) and stored at – 80° C until CPT analysis. Liver samples for CPT analysis were available for 39 cows. Sufficient liver for both composition and CPT analysis could not be obtained from all biopsies; only cows that had at least 1 prepartum sample and at least 2 postpartum samples were analyzed.
Activity of CPT in isolated mitochondria was assayed at 30° C using a modification of the method of Bremer (1981) as the formation of palmitoyl-[3H]-carnitine from palmitoyl-CoA and [3H]-L-carnitine (American Radiola-beled Chemicals Inc., St. Louis, MO). Briefly, mitochondria (~0.6 mg of protein/mL) were incubated for 3 min in reaction medium (75 mM KCl, 50 mM mannitol, 25 mM HEPES, 2 mM KCN, 0.2 mM EGTA, 1% BSA, 1 mM dithiothreitol, and 120 µM palmitoyl-CoA; pH 7.4), and then incubated for 6 min with carnitine solution (1,000 µM L-carnitine, ~1.5 µCi/mmol). The total volume was 1 mL. The reaction was terminated by the addition of 4 mL of 6% HClO4. Butanol was used to extract the palmitoyl-[3H]-carnitine formed by the reaction; 0.4 mL of the butanol phase was transferred to a scintillation vial, mixed with 10 mL of Scintisafe Econo 2 scintillation cocktail (Fisher Scientific, Atlanta, GA), and quantified by liquid scintillation spectroscopy (Beckman LS 6000IC, Beckman Instruments Inc., Fullerton, CA). The CPT activity was expressed as nanomoles of palmitoylcarnitine formed per minute per milligram of protein. The assay was validated by testing a range of substrate (palmitoyl-CoA and L-carnitine) concentrations, sample protein additions, and incubation times. Substrate concentrations chosen gave maximal activities and incubation conditions were within the linear range of the assay.
Statistical Analyses
Daily measurements such as DMI and milk production were reduced to weekly means for each cow before statistical analyses. Data were subjected to ANOVA for a randomized design with repeated measures by using the MIXED procedure of SAS (Littell et al., 1996; Release 8.2, SAS Institute Inc., Cary, NC). The model contained the effects of diet (C or F), prepartum intake of diet (A or R), the interaction of diet and prepartum intake, time, the interaction of diet with time, the interaction of prepartum intake with time, and the interaction of diet by prepartum intake with time. Cow nested within diet x prepartum intake was designated as a random term, and used as the error term to test the effect of dietary treatments. For repeated factors, several covariance structures were tested. The covariance structure that resulted in the Akaike’s information criterion closest to zero was used (Littell et al., 1996); in all cases this was the AR(1) option of SAS.
Separate ANOVA were conducted for intakes of DM and NEL during the prepartum period (– 60 d relative to expected calving to – 1 d relative to actual calving), prepartum transition period (3 wk before expected calving to – 1 d relative to actual calving), postpartum transition period (calving to 3 wk after calving), and postpartum period (calving to 105 d after calving). Separate ANOVA were conducted for BW and BCS data that were recorded during the prepartum and postpartum periods only. For concentrations of glucose, insulin, NEFA, cholesterol, BHBA, total protein, and urea N in plasma, separate ANOVA were conducted for the entire prepartum period, the transition period (21 d before the expected calving date to 21 d after the actual calving), and the entire postpartum period. The contents of total lipid, TG, and glycogen and the activity of CPT in liver during the prepartum and postpartum periods were combined for analysis; these data were also adjusted by analysis of covariance using the respective data obtained before dry-off at – 65 d relative to expected parturition. The incidences of health problems recorded during the study, including displaced abomasum, ketosis, mastitis, metritis, milk fever, and retained placenta, were subjected to 
2 analysis to determine whether incidences differed among treatment groups.
Because one cow was determined not to be pregnant immediately before expected parturition, it was eliminated from the experiment; therefore, n = 11 for CA, n = 12 for FA, n = 12 for CR, and n = 12 for FR. Least squares means were computed and are presented throughout. Significance was declared at P 0.05. Trends were considered present when 0.05 < P < 0.10.
|
RESULTS AND DISCUSSION
|
|---|
The C and F diets were offered to cows for 58.9 ± 4.8 d before parturition at both A and R intakes. Restricted-fed dry cows were offered amounts calculated to provide 80% of the NEL requirement based on BW at dry-off (NRC, 1989). Thus, R cows consumed an average of 7.4 kg of DM/d, which was significantly lower than that of A cows who consumed an average of 14.6 kg of DM/d during the dry period (Table 3; Figure 1A). Cows fed A and R consumed an average of 159 and 81%, respectively, of NRC (2001) requirements for NEL (Figure 1B). The substantial overconsumption of energy relative to requirements demonstrates that pregnant, nonlactating cows do not regulate intake to energy requirements, unless they are attempting to gain BCS (Friggens, 2003). Intakes of CP also were lower for cows fed R than for those fed A (153 vs. 77% of requirements; NRC, 1989). At mean DMI, BW, and BCS for the dry period (Table 3), estimated MP supply (NRC, 2001) was 1,196, 619, 1,246, and 600 g/d for CA, CR, FA, and FR groups, respectively. Consequently, A cows were provided with excess MP, whereas CR and FR were deficient by an estimated 22 and 39 g/d, respectively (NRC, 2001). Intakes of Ca and P exceeded requirements for both groups (not shown), although they were higher for A cows as expected. Diet composition (C vs. F) did not affect DMI (Table 3) or intakes of CP, Ca, or P (not shown).
View this table:
[in this window]
[in a new window]
|
Table 3. Production variables for cows fed control (C) or fat-supplemented (F) diets at either ad libitum (A) or restricted (R) intakes during the dry period
|
|
Dry cows fed A consumed an average of 2.2% of their BW as DM, confirming other observations that dry cows have the ability to consume large amounts of feed, even at high energy densities (Hayirli et al., 2003; Grummer et al., 2004). As parturition approached, DMI (Figure 2A; intake x day, P < 0.001) steadily decreased for dry cows fed A, which resulted in a 47% decrease in DMI from d – 21 until calving. Consequently, peripartal intakes of NEL (Figure 2B; intake x day, P < 0.001) as well as those of other nutrients (not shown) also decreased significantly during the last 3 wk before parturition for cows fed A. These data are consistent with previously reported decreases in DMI during the final days before calving, which have been reported to be on average about 30% from those earlier in the dry period (Grummer et al., 2004). However, the peripartal decreases in DMI for cows fed A in this study were greater than those reported by Grum et al. (1996), which were slightly less than 10%; this discrepancy may have been due in part to the poor quality forage included in the TMR and the lower DMI throughout the dry period in the Grum et al. (1996) study. Of interest is that R cows in our study, regardless of diet, maintained DMI and NEL intake as parturition approached and neither variable decreased until the day before calving (Figure 2). Cows fed R were in slight negative energy balance and lost BCS during the dry period; perhaps these factors overcame whatever inhibitory physiological signals were generated in ad libitumfed cows that caused decreased DMI. Higher energy diets seem to cause more pronounced decreases in DMI before calving (Grummer et al., 2004). The addition of the anionic mixture during the last 2 wk before parturition may have exacerbated the decrease in DMI for cows fed A if it was less palatable; however, the decrease began before the change of diet and an unpalatable diet might be expected to decrease DMI even in R cows.
View larger version (19K):
[in this window]
[in a new window]
|
Figure 2. Daily intakes of DM (panel A) and NEL (panel B; calculated according to NRC, 2001) during the transition period (21 d before to 21 d after parturition) for cows fed diets for ad libitum DMI ( ; n = 23) or restricted DMI ( ; n = 24) during the dry period. Panel A: Prepartum largest SEM = 0.6 kg/d; postpartum largest SEM = 1.0 kg/d. For prepartum data, effects of prepartum intake (P < 0.0001), day (P < 0.0001), and the interaction of prepartum intake x day (P < 0.0001) were significant. For postpartum data, effects of prepartum intake (P < 0.017) and day (P < 0.0001) were significant, but the interaction of prepartum intake x day was not significant (P = 0.19). Panel B: Prepartum largest SEM = 0.8 Mcal/d; postpartum largest SEM = 1.9 Mcal/d. For prepartum data, effects of prepartum intake (P < 0.0001), day (P < 0.0001), and the interaction of prepartum intake x day (P < 0.0001) were significant. For postpartum data, effects of prepartum intake (P < 0.035) and day (P < 0.0001) were significant, but the interaction of prepartum intake x day was not significant (P = 0.20).
|
|
Cows fed R during the prepartum period had greater DMI and NEL intake postpartum (Table 3; Figures 1 and 2) compared with cows that were fed A during the dry period. Others have found conflicting results on the effect of limit feeding during the dry period. Kunz et al. (1985) reported that cows fed restricted amounts during the final 70 d before calving of a diet formulated to meet maintenance requirements consumed significantly less DM and NEL during the dry period than ad libitumfed cows that consumed more NEL than required, but exhibited faster increases in DM and NEL intake immediately after calving. Studies published after our experiment was conducted (Tesfa et al., 1999; Holcomb et al., 2001; Agenäs et al., 2003) also found improvements in postpartum DMI by limiting intake during the dry period. In contrast, Boisclair et al. (1986) found no effect of prepartum feed restriction on postpartum DMI. The physiological mechanisms responsible for the increased DM and NEL intakes observed immediately postpartum following prepartum feed or energy restriction are not well understood, but may relate to the slightly lower BCS or chronic loss of BCS during the dry period for R cows. Mashek and Grummer (2003) suggested that changes in prepartum DMI might be more predictive of postpartum DMI than are absolute intakes prepartum. Our data support this theory; cows fed R during the dry period maintained DMI until the day before calving and had greater DMI postpartum, in contrast to the marked decreases in DMI as calving approached for cows fed A.
A secondary goal of this study was to determine the effects of BCS gain when cows were allowed to consume energy in excess of NRC (1989) recommendations during the dry period. Dry period diets were formulated so that a 650-kg dry cow could gain approximately 36 kg of BW in excess of fetal growth, or 0.6 BCS units. Restricted-fed cows were expected to lose BCS during the dry period. Gain of BCS by cows fed A was less than expected; cows fed CA and FA gained 0.14 and 0.20 BCS units from dry-off until approximately 3 wk before calving (Figure 3A). Grum et al. (1996) reported that dry cows fed a similar diet did not increase BCS; however, poor forage quality and lower DMI during the 60 d before calving limited energy intake. Cows fed R lost approximately 0.5 BCS unit from dry-off until approximately 3 wk before calving (Figure 3A). Interestingly, both A and R cows, regardless of energy source in the dry period diet, reached similar nadirs in postpartum BCS by 4 to 6 wk after calving (prepartum intake x week, P < 0.001). These observations support data from others showing that BCS converge postpartum despite different prepartum diets or BCS (Garnsworthy and Topps, 1982).
Regardless of the energy source used in the dry period diets, BW during the dry period tended (P = 0.08) to be greater for A cows (Table 3). Although BW increased for both A and R groups, cows fed A gained BW at a greater rate during the dry period until 3 wk before calving (Figure 3B; intake x week, P < 0.02). The plateau and slight loss of BW for A cows during the last 3 wk may reflect decreased DMI and gut fill as well as some mobilization of body tissue. These changes also may have been influenced by the greater number of twin births in A groups than in R groups (4 vs. 1; Table 4). Mean BW gains over the entire dry period were 56.2, 46.7, 27.9, and 18.9 kg for CA, FA, CR, and FR groups, respectively. Estimated BW gain attributable to fetal growth during the dry period is 35 to 40 kg (NRC, 2001). Because calf birth weights were not significantly different among treatment groups (Table 3), R cows must have mobilized more maternal tissues to support fetal development during the dry period. Peripartum patterns of BW loss were similar to those of BCS; cows fed A during the dry period lost significantly more BW during the 3 wk before and after calving, but both A and R groups reached similar nadirs in BW during the first 6 wk postpartum. Likewise, 175% (Rukkwamsuk et al., 1998) to 270% (Kunz et al., 1985) greater postpartum BW losses have been reported for cows that were overfed to promote BW gain prepartum compared with cows limit-fed to maintain or lose BW prepartum.
View this table:
[in this window]
[in a new window]
|
Table 4. Frequency1 of health conditions and twinning in cows fed control (C) or fat-supplemented (F) diets at ad libitum (A) or restricted (R) intakes during the dry period
|
|
Although mean milk production was about 2 kg/d higher during the first 105 d of lactation for cows fed R than for those fed A during the dry period (Table 3), the difference did not achieve statistical significance (P = 0.20). The higher postpartum DM and NEL intakes for cows previously restricted while dry would be expected to support higher milk production. Contents and yields of milk fat did not differ between treatments (Table 3) but decreased as lactation progressed (Figure 4; week, P < 0.001). However, a significant interaction of intake x week (P < 0.001) indicated that cows fed A while dry had higher milk fat contents during the immediate postpartum period (Figure 4B); milk fat content then declined to values similar to those of R cows. Higher milk fat content during the first 4 wk postpartum may have resulted from the lower DMI postpartum and greater losses of BCS postpartum for cows fed either diet at A intake while dry. Greater rates of body fat mobilization increase milk fat because of greater uptake and incorporation of NEFA (Palmquist et al., 1993). Similar trends were observed for milk fat yield (Figure 4C). Neither the content nor yield of CP in milk was different among treatment groups (Table 3).
The incidences of health problems during the experiment are reported in Table 4; 2 analysis suggested that only the incidence of metritis was affected by diet, with incidences lower (P < 0.01) for cows fed F than those fed C. Because the incidences of retained placenta and mastitis were not affected by treatment, it is unclear why metritis might have been less prevalent for cows fed F. The incidence of displaced abomasum was the only disorder affected significantly by intake, with fewer occurrences for cows fed either diet for R intake during the dry period (Table 4). Greater occurrence of displaced abomasum may reflect the lower DMI postpartum for cows fed A prepartum. The number of twin births tended to be greater for cows fed F, and was numerically greater for cows fed A than R. Deleting cows that calved with twins from the data set did not change direction or interpretation of results for production and metabolic variables; consequently, they were kept in the analysis. The small number of cows per treatment and the large amount of variation that accompanies peripartal health data demand that these data be interpreted with caution.
The lower prepartum intakes of DM and NEL for R cows resulted in significantly lower prepartum concentrations of glucose (Table 5, Figure 5) and insulin (Table 5) in plasma. Neither variable was affected by diet composition. The difference in glucose concentration between A and R cows confirms that energy status differed substantially between intake levels. In contrast, Kunz et al. (1985) reported that glucose concentrations were similar between groups of cows that were either overfed or limitfed to requirements during the dry period. Rukkwamsuk et al. (1998) also reported similar glucose concentrations at 1 wk before calving for dry cows offered feed for ad libitum or restricted intakes but did not measure glucose concentrations throughout the restriction period. Differences between these studies and the present study may have resulted because our cows were limit-fed below requirements rather than to meet requirements as in the other studies (Kunz et al., 1985; Rukkwamsuk et al., 1998).
View this table:
[in this window]
[in a new window]
|
Table 5. Concentrations of metabolites and insulin in plasma from cows fed control (C) or fat-supplemented (F) diets at either ad libitum (A) or restricted (R) intakes during the dry period
|
|
View larger version (18K):
[in this window]
[in a new window]
|
Figure 5. Concentrations of glucose (panel A), NEFA (panel B), and BHBA (panel C) in plasma from cows fed control (diamonds) or fat-supplemented (triangles) diets for ad libitum DMI (filled symbols and solid lines) or restricted DMI (open symbols and broken lines) during the dry period. Panel A: Prepartum largest SEM = 1.9 mg/dL; postpartum largest SEM = 1.9 mg/dL. For prepartum data, effects of prepartum intake (P < 0.0002) and week (P < 0.0001) were significant, and the interaction of prepartum intake x week (P < 0.06) approached significance. For postpartum data, the effects of week (P < 0.0001) and the interaction of prepartum intake x diet x week (P < 0.041) were significant. Panel B: Prepartum largest SEM = 0.066 mM; postpartum largest SEM = 0.098 mM. For prepartum data, the effects of prepartum intake (P < 0.0001), week (P < 0.0001), and the interaction of prepartum intake x week (P < 0.039) were significant. For postpartum data, effects of prepartum intake (P < 0.040) and week (P < 0.0001) were significant, and the interaction of prepartum intake x diet x week (P = 0.073) approached significance. Panel C: Prepartum largest SEM = 0.5 mg/dL; postpartum largest SEM = 1.4 mg/dL. For prepartum data, the effects of week (P < 0.0001), the interaction of prepartum intake x week (P < 0.004), and the interaction of prepartum diet x week (P < 0.010) were significant. For postpartum data, effects of prepartum intake (P < 0.045), week (P < 0.0001), the interaction of intake x week (P < 0.0001), and the interaction of prepartum intake x diet x week (P = 0.040) were significant.
|
|
As calving approached, concentrations of glucose (Figures 5 and 6) and insulin (Figure 6) in plasma decreased for all treatment groups. Grum et al. (1996) reported similar patterns for peripartum glucose concentrations; others have reported transient increases in glucose and insulin concentrations at calving followed by marked decreases after parturition (Vazquez-Añon et al., 1994; Kunz et al., 1985; Rukkwamsuk et al., 1998). Differences among studies may reflect differences in peripartal DMI. Regardless, the significant decreases in DM and NEL intakes during the prepartum transition period by cows fed A during the dry period in our study probably resulted in the more rapid decreases in plasma glucose (intake x day, P < 0.02) and insulin (intake x day, P < 0.0001; Figure 6). After calving, glucose and insulin concentrations were not different among treatments (Table 5), although the interaction of prepartum intake x diet x week was significant (Figure 5A). The interaction resulted mainly from the sharp drop of plasma glucose for cows in the FA group at wk 2; the reason for this decrease is unknown.
As expected, R cows had significantly higher concentrations of NEFA in plasma throughout the dry period than did A cows (Table 5, Figure 5B), in which NEFA began to increase only during the last 2 wk before calving. Similar trends were reported by Kunz et al. (1985) for cows that were limited to maintenance intakes during the final 70 d of gestation and by Grum et al. (1996) for cows fed a high-roughage, fat-supplemented diet relative to groups that consumed greater than required amounts of NEL. The concomitantly lower concentrations of glucose and insulin in plasma resulting from feed restriction throughout the dry period would be associated with moderately higher rates of adipose tissue TG mobilization to support energy needs, and was reflected in the sustained loss of BCS throughout the dry period (Figure 3). Concentrations of BHBA in plasma (Table 5, Figure 5C) did not differ significantly among treatment groups during the dry period; the difference between R and A did not achieve statistical significance (P < 0.14). During the last week before calving, NEFA concentration increased more for cows fed A during the dry period (Figure 6C). In contrast to our data, dry cows fed to 67 or 100% of energy requirements for a short period (from d 30 to 10 before expected calving date) exhibited no differences in blood concentrations of NEFA, BHBA, or glucose (Lotan et al., 1988).
Mean postpartum concentrations of NEFA and BHBA were greater for cows fed A during the dry period (Table 5). The interaction of prepartum intake x day also was significant for peripartal concentrations of NEFA (Figure 6C) and BHBA (Figure 7). These interactions indicate that both variables were greater after calving for cows that were fed A while dry, as a result of the lower postpartum DMI and greater postpartum BCS loss for A cows. The 3-way interactions of prepartum intake x diet x week (or day) were significant for BHBA (Figures 5C and 7). Cows that were fed FA while dry had delayed but more pronounced increases in BHBA postpartum than cows fed CA or either diet at restricted intake. The larger increases of BHBA during wk 2 for cows fed FA corresponded to the lower glucose concentration for that group (Figure 5A). Cows that were restricted-fed either diet during the dry period maintained lower plasma BHBA after parturition; cows fed FA tended to have higher BHBA postpartum than those fed FR (Figure 7). Concentrations of BHBA converged for all groups by about wk 5 postpartum (Figure 5C), whereas NEFA concentrations remained somewhat elevated for cows previously fed CA through at least wk 8 postpartum (Figure 5B).
View larger version (13K):
[in this window]
[in a new window]
|
Figure 7. Concentration of BHBA in plasma during the transition period (21 d before to 21 d after parturition) from cows fed control (diamonds) or fat-supplemented (triangles) diets for ad libitum DMI (filled symbols) or restricted DMI (open symbols) during the dry period. Largest SEM = 1.9 mg/dL. The effects of prepartum intake (P < 0.020), day (P < 0.0001), the interaction of prepartum intake x day (P < 0.0001), and the interaction of prepartum intake x diet x day (P = 0.050) were significant.
|
|
Supplementing fat (FA, FR) in the dry period diet increased prepartum concentrations of cholesterol in plasma regardless of intake (Table 5, Figure 8A). Increasing the amounts of LCFA reaching the small intestine of dairy cows increases cholesterol concentrations in plasma (Bremmer et al., 1998; Douglas et al., 2004). Plasma cholesterol decreased as calving approached (week, P < 0.0001). A significant interaction of diet x week (P < 0.0001) indicated that cholesterol concentrations decreased later in the dry period for cows fed F than for those fed C, although the rate of decrease was greater for cows fed F (Figure 8A). Surprisingly, cows fed diets at R intake had greater concentrations of cholesterol in plasma than cows fed A (Table 5). Most cholesterol synthesis in ruminants is believed to occur in the small intestinal epithelium to transport dietary lipid (Nestel et al., 1978); therefore, lower concentrations in feed-restricted cows might be expected because of lower DMI. Reasons for the increased cholesterol concentrations in the R cows are not known, but it is possible that cholesterol clearance or lipoprotein metabolism was altered in these cows. After parturition, cows that were previously fed R while dry had greater cholesterol concentrations throughout the first 15 wk of lactation (Table 5), which probably resulted from the higher postpartum DMI. Prepartum diet did not affect postpartum cholesterol concentration.
Despite the lower CP intakes for R cows during the dry period, no differences were observed for the concentration of total protein in plasma (Table 5). Plasma total protein is decreased during protein malnutrition (Swenson, 1993). Total protein steadily decreased (week, P < 0.0001) during the last 4 wk of gestation regardless of dietary treatment, but concentrations returned to those observed during the early dry period by 3 to 4 wk postpartum (data not shown). Similar patterns for plasma total protein during the transition period following energy restriction during the dry period were reported by Kunz et al. (1985). Plasma urea N concentrations did not differ between A and R groups during the dry period (Table 5). Plasma concentrations of urea N steadily decreased as calving approached (week, P < 0.0001); a significant interaction of diet x week (P < 0.02) indicated that cows fed F during the dry period had higher urea N early in the dry period but lower urea N by calving (Figure 8B). The interaction of prepartum intake x diet was significant for urea N postpartum (Table 5), although we cannot explain such a response. Plasma concentrations of urea N increased as lactation progressed (week, P < 0.0001; Figure 8B).
Increased NEFA concentrations in blood have been associated with increased risk for development of fatty liver and, subsequently, for development of other peripartum metabolic disorders including ketosis (Grummer, 1993; Bobe et al., 2004). Interestingly, cows that were restricted-fed while dry had significantly lower concentrations of total lipid (intake x day, P < 0.01) and TG (intake x day, P < 0.01) in the liver at 1 and 21 d postpartum (Figure 9) despite chronically elevated concentrations of NEFA throughout the dry period. The NEFA concentrations in R cows during the dry period were not high enough to cause hepatic TG accumulation (Drackley et al., 2001). Both A and R cows had marked increases of NEFA immediately before parturition. These data agree with previous observations that liver lipid accumulation occurs in the presence of increased plasma concentrations of NEFA at calving (Gerloff et al., 1986; Skaar et al., 1989; Vazquez-Añon et al., 1994). However, TG contents at d 1 for A cows were approximately double those of R cows. Peak NEFA concentrations (at d 2; Figure 6C) were similar between A and R cows, but NEFA remained elevated for longer in A cows. Grum et al. (1996) suggested that chronically increased NEFA and low insulin during the dry period led to increased oxidative enzymes in liver and decreased esterification activity for NEFA, which could lead to decreased TG accumulation.
Concentrations of TG in liver tissue from cows previously fed F tended (P < 0.10) to be lower than from cows fed C; at d 1 postpartum TG concentrations were 6.6, 5.0, 3.6, and 2.8% of wet weight for cows previously fed CA, FA, CR, and FR, respectively. Others have reported increases in plasma concentrations of NEFA with dietary fat supplementation prepartum and suggested a greater risk for development of fatty liver (Skaar et al., 1989). However, the present data and our previously published data (Douglas et al., 2004), as well as data from others (Doepel et al., 2002), indicate no increased hepatic lipid accumulation due to dietary fat. Effects of prepartum fat supplementation on postpartum intake and production have been unremarkable (Skaar et al., 1989; Salfer et al., 1995; Douglas et al., 2004). Doepel et al. (2002) compared low-energy diets with diets in which energy density was increased by addition of 2.2% tallow and decreased forage-to-concentrate ratio during the last 21 d before parturition. Although they reported improvements in postpartum DMI, energy balance, and liver TG content with higher energy diets, the small number of cows per treatment and the fact that cows were fed differently during late lactation and the early dry period depending on BCS make comparison with our results difficult.
Liver glycogen contents (Figure 9C) were not statistically different among treatment groups. Concentrations of glycogen in liver decreased dramatically at d 1 postpartum regardless of prepartum intake or source of dietary energy and then increased to concentrations observed at dry-off by d 65 (day, P < 0.0001). However, a significant interaction of prepartum intake x day indicated that cows fed A had higher hepatic concentrations of glycogen than did R cows at 21 d before calving (Figure 9C). Glycogen concentrations in liver were similar between intake groups at d 1 postpartum but then were slightly greater for R cows than A cows by d 21, which may reflect the greater postpartum DMI by R cows. Similar peripartal changes in liver glycogen concentrations were reported by Grum et al. (1996).
We interpret our data to indicate that a lower overall plane of nutrition prepartum compared with overfeeding results in markedly less accumulation of total lipid and TG in the liver after parturition. Diet (i.e., the addition of fat in replacement for NFC) had a smaller impact on peripartal TG accumulation in liver. Together with results of Douglas et al. (2004), our data indicate that the marked decrease in hepatic lipid accumulation observed in the study of Grum et al. (1996) was predominantly caused by the decreased energy and nutrient intakes prepartum for cows fed the high-fat, low-NFC diet in that study rather than by fat supplementation per se. In that study, the fat-supplemented group had lower DMI and lost BCS; in contrast, cows fed the lower energy density control diet consumed near requirements and cows fed a higher energy diet containing more grain consumed > 145% of energy requirements and yet both accumulated more liver TG than the fat-supplemented cows (Grum et al., 1996).
The mechanisms responsible for the reduction of peripartal liver lipid accumulation observed with lower energy intake during the dry period are not fully known or understood. Rukkwamsuk et al. (1998) and Drackley (1999) concluded that the decreased hepatic TG accumulation following feed restriction during the dry period was due to alterations in metabolism that better adapted the liver to contend with the marked peripartum increases in blood NEFA. Grum et al. (1996) reported that cows fed a fat-supplemented diet through-out the dry period had lower NEL intake, lost BCS, and had higher concentrations of NEFA in plasma than cows fed isoenergetic diets without fat. Consequently, those cows were physiologically similar to the cows that were restricted-fed in our study. Cows fed the fat-supplemented diet in the study by Grum et al. (1996) had numerically greater total ß-oxidative capacity, significantly increased rates of peroxisomal ß-oxidation, and a higher ratio of peroxisomal to total ß-oxidation in liver tissue at d 1 postpartum. Moreover, in vitro rates of palmitate esterification were significantly lower in liver tissue obtained from cows previously fed the fat-supplemented diet compared with liver from controls or cows fed higher grain diets that consumed more energy (Grum et al., 1996). A subsequent study also provided evidence that peroxisomal ß-oxidation may augment extant, albeit sometimes inadequate, oxidative pathways for lipid disposal in liver during the peripartal period (Grum et al., 2002). Therefore, increases in mitochondrial and peroxisomal ß-oxidation, as well as decreases in esterification of LCFA in the liver, may have been responsible for the markedly lower hepatic accumulation of TG in cows restricted-fed either diet in the present study compared with the overfed A cows.
To determine the potential role of adaptations in mitochondrial ß-oxidation on hepatic lipid accumulation, we measured mitochondrial CPT activity. No interaction of diet with amount fed was detected (P = 0.84); only main effects are presented. Cows fed C during the dry period tended (P = 0.10) to have higher CPT activity than cows fed F during the dry period (Figure 10A). Due to the lack of dietary treatment effects on concentrations of glucose, insulin, and NEFA in plasma and on hepatic contents of total lipid and TG, it is difficult to suggest a mechanism for higher CPT activity in cows fed C during the dry period.
View larger version (27K):
[in this window]
[in a new window]
|
Figure 10. Means and standard errors for main effects of diet (panel A) or prepartum intake (panel B) on carnitine palmitoyltransferase (CPT) activity in isolated mitochondria from liver of multiparous Holstein cows. The mean pretreatment (pretrt) value at d – 65 also is shown. Panel A: Effects in the model included diet (P = 0.10), day (P = 0.001), and the interaction of diet x day (P = 0.90). Panel B: Effects in the model included prepartum intake (P = 0.96), day (P = 0.001), and the interaction of prepartum intake x day (P = 0.08).
|
|
Prepartum intake did not affect CPT activity (P = 0.96) when all time points were included in the model, but a tendency (P = 0.08) for an interaction of prepartum intake x time was detected (Figure 10B). When CPT activity was analyzed by individual day, cows fed R during the dry period tended to have higher CPT activity at – 21 d (P = 0.06) and 1 d (P = 0.05) and lower CPT activity at 65 d (P = 0.06) than cows fed A during the dry period. Activity of CPT did not differ between treatments at 21 d (P = 0.50). Cows fed R prepartum had a greater change in postpartum CPT activity. During the prepartum period, cows fed R had higher NEFA, lower glucose, and lower insulin concentrations in plasma than cows fed A. The plasma data are consistent with the state of negative energy balance for R cows and an increase in CPT activity as observed in other species (Bremer, 1981; Park et al., 1995; Kerner and Hoppel, 2000). During the postpartum period, concentrations of NEFA in plasma were higher for cows fed A during the dry period (Table 5, Figure 6C), indicating a more negative energy balance that may have increased CPT activity at 65 d. Thus, increased ß-oxidative capacity may have decreased LCFA available for esterification in the liver and therefore contributed to lower hepatic lipid and TG accumulation for cows fed R. A relatively modest increase in activity of the rate-limiting enzyme CPT combined with potential changes in hepatic concentration of its inhibitor malonyl-CoA (Dann and Drackley, 2005) and modest changes in NEFA flux could result in substantial changes in hepatic TG deposition (Drackley et al., 2001).
Day relative to parturition affected CPT activity (P < 0.001; Figure 10). Similar to the peak in plasma NEFA, CPT activity was higher at d 1 than at d – 21 (11.33 vs. 7.92 nmol of palmitoylcarnitine produced/min per mg of protein; P = 0.01), higher at d 1 than at d 65 (11.33 vs. 7.77; P < 0.001), and similar between d 1 and d 21 (11.33 vs. 10.12; P = 0.19). The CPT activity followed a pattern generally similar to that of total lipid and TG contents in liver (Figure 9). Day relative to parturition had a greater effect on fatty acid metabolism and CPT activity than did diet composition or intake level. Aiello et al. (1984) found that CPT activity was highest at 30 d postpartum and decreased as lactation progressed.
Insulin has been shown to increase TG accumulation in rodent liver (Gibbons et al., 2004). Effects of insulin on esterification of LCFA in ruminant liver are less clear (Emmison et al., 1991; Drackley et al., 2001). Insulin suppresses hepatic ß-oxidation of LCFA in dairy cows (Andersen et al., 2002), possibly by effects on CPT (Jesse et al., 1986). Therefore, the higher concentrations of insulin in plasma during the dry period for A cows may have suppressed ß-oxidation and increased esterification of LCFA with the resultant increase in hepatic TG accumulation at calving. Lower insulin concentrations resulting from limiting energy intake have been associated with decreased peripartum liver TG accumulation in other studies (Grum et al., 1996; Rukkwamsuk et al., 1998).
Duration of energy restriction as well as physiological state may be important in eliciting metabolic adaptations in peripartal cows. Grum et al. (1994) reported that 7 d of feed deprivation in nonlactating, nonpregnant cows did not alter peroxisomal ß-oxidation in liver despite large differences in plasma NEFA concentrations. The present data support the hypothesis that longer-term energy restriction during the dry period may be necessary to induce peripartum alterations in hepatic lipid metabolism. Whether feeding cows to meet but not exceed NEL requirements would elicit similar changes relative to overfed cows cannot be determined from our data. Considerable evidence indicates that restricted or limited feeding throughout the dry period may invoke favorable metabolic and intake responses (Kunz et al., 1985; Tesfa et al., 1999; Holcomb et al., 2001; Agenäs et al., 2003; Holtenius et al., 2003), although not all reports agree (Boisclair et al., 1986; Lotan et al., 1988). VandeHaar et al. (1999) reported lower hepatic TG after parturition for cows that were fed diets of greater energy density during the last 3 wk of the dry period (i.e., the close-up period) compared with controls fed lower-energy diets. Effects of this common two-phase approach to dry cow management on hepatic LCFA metabolism cannot be predicted from the present data.
|
CONCLUSIONS
|
|---|
The primary objective of this experiment was to determine if decreases in hepatic concentrations of total lipid and TG in liver observed in cows fed a high roughage, fat-supplemented diet in the study of Grum et al. (1996) were due to fat supplementation or to the lower energy and nutrient intakes. We conclude from our data that decreased hepatic lipid accumulation in the earlier study was more likely a result of prepartum energy and nutrient restriction than the source of energy included in the dry period diet. Feeding a fat-supplemented diet during the dry period tended to reduce peripartal liver lipid accumulation, but effects were smaller than the effect of plane of nutrition. Limiting energy intake for the entire dry period may enhance hepatic capacity for ß-oxidation, as suggested by increased CPT activity, and decrease esterification of LCFA, thereby leading to less accumulation of lipid and TG in the liver during the immediate postpartum period.
Unlike cows in the study of Grum et al. (1996), cows allowed to overconsume energy during the dry period in our study gained BCS until about 3 wk before parturition. Although these cows did not become overconditioned, they had marked decreases in DMI before calving, and had higher plasma concentrations of NEFA and BHBA, higher hepatic concentrations of total lipid and TG, and lower DMI postpartum than cows that were restricted-fed while dry. We suggest from our data that allowing ad libitum access to diets containing moderate to high energy densities throughout the entire dry period to allow BW gain could be detrimental to peripartum health and postpartum performance even when cows do not become overconditioned. Moreover, there were no advantages of feeding to allow gain of BCS during the dry period, because subsequent milk production was not improved and cows reached similar nadirs in BCS by 5 wk postpartum.
|
ACKNOWLEDGEMENTS
|
|---|
The authors extend appreciation to G. C. McCoy and employees of the University of Illinois Dairy Research and Teaching Unit for provision and care of cows. Soybean hulls were kindly donated by Archer Daniels Midland Co. (Decatur, IL).
|
FOOTNOTES
|
|---|
1 This project was funded by a grant from the Fats and Proteins Research Foundation and by USDA Section 1433 Animal Health and Disease Funds appropriated to the Illinois Agricultural Experiment Station. Heather M. Dann was supported by a Jonathan Baldwin Turner graduate fellowship from the College of Agricultural, Consumer and Environmental Sciences, University of Illinois. 
2 Current address: CPO 1862, Berea College, Berea, KY 40404.
3 Current address: Department of Animal Science, Cornell University, Ithaca, NY 14853.
4 Current address: Akey, PO Box 5002, Lewisburg, OH 45338.
5 Current address: William H. Miner Agricultural Research Institute, Chazy, NY 12921.
Received for publication July 24, 2005.
Accepted for publication November 28, 2005.
|
REFERENCES
|
|---|
Agenäs, S., E. Burstedt, and K. Holtenius. 2003. Effects of feeding intensity during the dry period. 1. Feed intake, bodyweight, and milk production. J. Dairy Sci. 86:870–882.[Abstract/Free Full Text]
Aiello, R. J., T. M. Kenna, and J. H. Herbein. 1984. Hepatic gluconeogenic and ketogenic interrelationships in the lactating cow. J. Dairy Sci. 67:1707–1715.[Abstract/Free Full Text]
Andersen, J. B., D. G. Mashek, T. Larsen, M. O. Nielsen, and K. L. Ingvartsen. 2002. Effects of hyperinsulinaemia under euglycaemic condition on liver fat metabolism in dairy cows in early and mid-lactation. J. Vet. Med. A 49:65–71.
Association of Official Analytical Chemists.1995. Official Methods of Analysis. 16th ed. AOAC, Arlington, VA.
Bobe, G., J. W. Young, and D. C. Beitz. 2004. Pathology, etiology, prevention, and treatment of fatty liver in dairy cows. J. Dairy Sci. 87:3105–3124.[Abstract/Free Full Text]
Boisclair, Y., D. G. Grieve, J. B. Stone, O. B. Allen, and G. K. Macleod. 1986. Effect of prepartum energy, body condition, and sodium bicarbonate on production of cows in early lactation. J. Dairy Sci. 69:2636–2647.[Abstract/Free Full Text]
Bremer, J. 1981. The effect of fasting on the activity of liver carnitine palmitoyltransferase and its inhibition by malonyl-CoA. Biochim. Biophys. Acta 665:628–631.[Medline]
Bremmer, D. R., L. D. Ruppert, J. H. Clark, and J. K. Drackley. 1998. Effects of chain length and unsaturation of fatty acid mixtures infused into the abomasum of lactating dairy cows. J. Dairy Sci. 81:176–188.
TOP
ABSTRACT
INTRODUCTION
; n = 23) or fat-supplemented (
; n = 24) diets during the dry period. Panel A: Prepartum largest SEM = 5.3 mg/dL; postpartum largest SEM = 7.6 mg/dL. For prepartum data, effects of prepartum intake, prepartum diet, week, and the interaction of prepartum diet x week were significant (P < 0.0001). For postpartum data, effects of prepartum intake (P < 0.004) and week (P < 0.0001) were significant. Panel B: Prepartum largest SEM = 0.76 mg/dL; postpartum largest SEM = 0.74 mg/dL. For prepartum data, the effects of week (P < 0.0001) and the interaction of prepartum diet x week (P < 0.030) were significant. For postpartum data, only the effect of week (P < 0.0001) was significant.