Doi:10.1016/j.theriogenology.2004.03.002

Increasing feed intake in late gestation does not affect plasma progesterone concentration in the sow aSchool of Biology, University of Leeds, Leeds LS2 9JT, UK bDepartment of Agriculture, Food and Nutritional Science, University of Alberta, Edmonton, Alta., Canada Received 10 February 2003; received in revised form 20 February 2004; accepted 14 March 2004 Rate of decline in plasma progesterone concentration may influence the success of lactogenesis in the sow. The aim of this experiment was to investigate whether progesterone concentration and rate ofdecline of progesterone in the periparturient sow could be manipulated by changing her feeding level.
Forty-two sows received either 1.15 or 2 times maintenance energy daily from day 100 of gestation upuntil and including the day of farrowing. Blood samples were taken on days 98 (pre-treatmentbaseline) and 109 of gestation, during farrowing, 6 h after farrowing and at 09:00 h for the 3 daysfollowing farrowing. Plasma progesterone concentration was determined and progesterone half-lifewas calculated for each sow. High intake feeding had no effect on plasma progesterone concentrationat any time of sampling. Progesterone half-life averaged 41:2 Æ 3:81 h and did not differ betweentreatments. There was no relationship between progesterone concentration, or half-life, and litterweight gain, although there was a weak correlation between decline in progesterone in the first 6 hafter birth and piglet growth rate from birth to 6 days of age (R2 ¼ 0:109, P < 0:05). It was concludedthat increasing feed intake in late gestation cannot be used to increase progesterone clearance rate andhasten the onset of lactogenesis in sows.
# 2004 Elsevier Inc. All rights reserved.
Keywords: Progesterone; Sow; Food intake; Late gestation There are two periods in the reproductive cycle of the sow when it might be advantageous to manipulate progesterone concentrations. The first of these is immediatelyfollowing conception in order to maximise embryo survival The second period is in late * Corresponding author. Tel.: þ44-113-3433054; fax: þ44-113-3433144.
E-mail address: h.m.miller@leeds.ac.uk (H.M. Miller).
0093-691X/$ – see front matter # 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.theriogenology.2004.03.002 H.M. Miller et al. / Theriogenology 62 (2004) 1618–1626 pregnancy as the sow prepares for farrowing and lactation. The sow has maintained highprogesterone concentrations throughout pregnancy. Progesterone is lipophilic; cyclinggilts have been shown to have 10 times greater concentration of progesterone in fat than inblood . Therefore, sows approaching the end of pregnancy are likely to have aconsiderable quantity of progesterone stored in their body fat. Progesterone withdrawalis necessary to allow parturition and has also been implicated in lactogenesis in the sow.
de Passille´ et al. found that there was a negative correlation between sow plasmaprogesterone concentrations during the first 48 h after farrowing and piglet growth rates inthe first 3 days of life.
It has been clearly demonstrated in the progesterone infused ovariectomised gilt that increasing feed intake level increases metabolic clearance rate of progesterone . Howwell does this model reflect the impact of feeding level in the intact female pig? There isevidence of an inverse relationship between feed intake and plasma progesterone con-centrations during the first half of pregnancy however, the effect of increasing feedintake as a means of increasing the rate of decline of progesterone prior to parturition hasnot been reported. The aim of this experiment was to investigate whether we couldmanipulate progesterone concentration and rate of decline of progesterone in the peri-parturient sow simply by changing her feeding level. We hypothesised that increasing feedintake in late gestation would reduce progesterone concentration in the peri-parturient sow.
All procedures used in this experiment were approved by the University of Alberta Animal Care Committee to ensure adherence to Canadian Council of Animal Care Guidelines.
Forty-six Camborough sows (Pig Improvement (Canada) Ltd., Acme, AB, Canada), parities 1 to 3, were fed 1:15 Â maintenance energy levels until day 100 of gestation(normal). From day 100 of gestation until farrowing the sows were randomly allocatedwithin parity either to remain on normal feed intake level (1:15 Â maintenance energy—Nsows), or to receive a high feed intake (twice maintenance energy—H sows). The high levelof intake was set at twice maintenance rather than twice the normal intake becausepreliminary trials experienced difficulties in getting sows to consistently eat the greateramount of feed. There were 23 H sows (9 Â parity 1, 10 Â parity 2 and 4 Â parity 3) and 23N sows (10 Â parity 1, 8 Â parity 2 and 5 Â parity 3), unfortunately four of the parity 1 Hsows refused to eat their allocated ration in the last week of gestation and were removedfrom the trial. Reported results are therefore based on the remaining 42 sows. All sowswere weighed on day 98 of gestation and within 24 h of farrowing. Backfat thickness wasmeasured at these same times using an ultrasonic probe (Scanoprobe II, Scanco, Ithaca,NY) at the last rib and 65 mm from the midline (P2).
Maintenance requirement was assumed to be 460 kJ DE/kg BW0.75 . The daily ration was fed once daily at 8 a.m. The gestation diet contained 12.6 MJ DE/kg, 13.7% crudeprotein and 0.56% lysine. Sows on each treatment had similar weight and fatness at day 98of gestation, averaging 199:7 Æ 4:93 kg live weight and 17:7 Æ 0:46 mm P2 backfat.
During gestation the sows were housed in individual stalls at a room temperature of20 Æ 2 8C. On day 109 of gestation the sows were moved into individual farrowing crates H.M. Miller et al. / Theriogenology 62 (2004) 1618–1626 in the farrowing room and from this time until the end of lactation they received anappropriate amount of the lactation diet (13.7 MJ DE/kg; 15.4% crude protein: 0.74%lysine). Farrowing stalls were totally slatted and had covered creep areas containing heatlamps and overlays. Room temperature was maintained at 18 Æ 2 8C. After farrowing allsows were fed ad libitum throughout a 26 (Æ0.3) day lactation. Feed intake was recordeddaily. Sows had access to water at all times.
Litter size was standardised across treatments within 24 h of parturition. On the day of birth, piglets received an iron supplement, their teeth and tails were clipped and their earswere notched. Male piglets were castrated at 7 days of age. Piglets were weighed within 6 hof birth and then at 1, 2, 6 and 20 days of age and at weaning. Piglet weight change in thefirst 2 days of life was used as an indicator of the onset of lactogenesis. Piglets had access tocreep feed from 20 days of age and to water at all times. Intake of creep feed was notrecorded.
Blood samples were taken on day 98–2 h after feeding; day 109—immediately before and 2 h after feeding; during farrowing, 6 h later and at 9 a.m. on the next 3 days. Bloodsamples were taken by ear vein puncture without sow restraint. All blood samples werecentrifuged at 1500 Â g for 15 min at 4 8C within 20 min of sampling and plasma wasstored at À30 8C until analysis.
Plasma samples were analysed for progesterone by radioimmunoassay as described by Beltranena et al. Sensitivity of the assay was 0.02 ng/100 mL. No significant deviationfrom parallelism was apparent from assaying 100, 50 and 25 mL of a standard plasma pool.
The intra- and inter-assay coefficients of variation for the progesterone assays were 4.1 and5.2%, respectively.
The disappearance rate constant (k is the fraction of total hormone that disappears per unit time) of progesterone after farrowing was determined by regression of the naturallogarithm of plasma progesterone concentrations from during farrowing until 3 days post-partum against time. The resulting regression fits the equation X ¼ A eÀkt, where k is theslope of the curve. Only sows for which this regression was significant at the 5% level wereincluded in the analysis. Progesterone half-life was calculated from the following equation: Data for pre-treatment plasma progesterone concentration, disappearance rate constant, half-life, piglet number, BW and growth rate were analysed using analyses of variance.
Sources of variation were treatment (high or normal feed intake during the last 2 weeks ofgestation), parity (p ¼ 3), interaction between feed intake and parity and sow within feed H.M. Miller et al. / Theriogenology 62 (2004) 1618–1626 intake by parity (error term). Preliminary analyses indicated no significant feed intake byparity interactions and therefore only main effect least square means are presented.
Comparisons among least square means were made using Fisher’s protected leastsignificant difference All computations were made using MANOVA operation ofSystat Data for plasma progesterone concentrations during the treatment period and atfarrowing were analysed according to the procedure described above plus a covariate ofpre-treatment plasma progesterone concentration. Comparison between progesteroneconcentration of post-prandial samples before treatment (day 98) and during treatment(day 109) was computed as a split-plot addition to the model given above for pre-treatmentprogesterone concentration. Likewise comparison between progesterone concentration ofpre-prandial and post-prandial samples during treatment (day 109) were also computed as asplit-plot addition to the model given above for pre-treatment progesterone concentration.
There were no significant interactions between treatment and parity for any of the From day 100 of gestation until farrowing mean daily intakes of H and N sows were 3:9 Æ 0:09 kg/day and 2:4 Æ 0:07 kg/day, respectively (P < 0:001). After farrowing therewas no difference in feed intake between the two groups, mean feed intake over the wholelactation was 6:5 Æ 0:26. There was no difference in P2 backfat thickness at farrowingbetween the treatments which averaged 16:8 Æ 0:95 mm for H sows and 16:6 Æ 0:85 mmfor N sows.
There was no clinical incidence of agalactia or mastitis in this experiment.
There was no difference between plasma progesterone concentrations of H and N sows at any stage measured in late gestation or early lactation, nor in the decline in progesteronein the first 6 h after farrowing (see and Pre-treatment plasma progesterone Table 1Plasma progesterone concentrations during late gestation and early lactation in sows fed high or normal intakesin late gestation During treatment (day 109)—pre-prandial 1 Pre-treatment value used as covariate.
H.M. Miller et al. / Theriogenology 62 (2004) 1618–1626 Fig. 1. Plasma progesterone concentrations in late gestation and early lactation for sows which received normalor high levels of feed intake in late gestation. X-axis: time from farrowing (days). Y-axis: progesteroneconcentration (ng/mL).
concentrations on day 98 of gestation varied considerably between sows and werecorrelated to sow liveweight (R2 ¼ 0:221, P < 0:001). Plasma progesterone on day 98of gestation was not related to the number of piglets in the litter. Parity 1 sows had lowerplasma progesterone concentrations then sows of parities 2 and 3 on day 98 of gestation(P < 0:05, but thereafter there were no differences in progesterone concentrationbetween sows of different parities.
Plasma progesterone concentrations of high intake sows did not decline between days 98 and 109 of gestation in response to increased feeding and were not different from those ofnormal intake sows. On day 109 of gestation, progesterone concentrations were similarbefore and after feeding. Progesterone concentrations at farrowing were best predicted Table 2Plasma progesterone concentrations during late gestation and early lactation in sows of parities 1, 2 and 3 Values with different superscripts within rows are significantly different.
1 Pre-treatment value used as covariate.
H.M. Miller et al. / Theriogenology 62 (2004) 1618–1626 from the inherent progesterone concentration of the sow (on day 98 of gestation) and thetotal number of piglets born: P4f ¼ À0:57 Æ 2:05 þ 0:23 Æ 0:086 P4d98 þ 0:28 Æ 0:122 LSb where P4f is sow plasma progesterone concentration at farrowing (ng/mL). P4d98 is sowplasma progesterone concentration on day 98 of gestation (ng/mL). LSb is total number ofpiglets born.
Disappearance rate constants could only be calculated for 34 of the 42 sows and were not different between treatments. Progesterone half-life averaged 41:2 Æ 3:81 h. Regressionanalysis identified a positive relationship between disappearance rate constant (k) andplasma progesterone at farrowing but a negative relationship to litter size at birth and to sowbackfat thickness at farrowing. Therefore, half-life was shorter when progesterone atfarrowing was higher, but became longer as the number of piglets in the litter, or the fatnessof the sow at farrowing, increased; k ¼ 0:047 Æ 0:010 þ 0:002 Æ 0:001 P4f À 0:002 Æ 0:001 Bf À 0:001 Æ 0:001 LSb where k is progesterone disappearance rate constant. Bf is sow P2 backfat thickness atfarrowing.
There was no relationship between progesterone concentrations during and after farrowing, nor the rate of decline of progesterone, with piglet weight gain in the first 2days of life. There was a significant although weak correlation between decline inprogesterone in the first 6 h after birth and piglet growth rate from birth to 6 days ofage (R2 ¼ 0:109, P < 0:05). The disappearance rate constant was not correlated to anymeasure of piglet growth.
Piglet birthweight was not affected by feed intake in late gestation and there were no significant differences in any other aspect of piglet performance between the two treatment Table 3Piglet performance data for sows fed either high or normal feeding levels in late gestation Daily piglet liveweight gain to 20 days (kg/day) Daily piglet liveweight gain to weaning (kg/day) H.M. Miller et al. / Theriogenology 62 (2004) 1618–1626 groups (see ). Sows produced an average of 9:9 Æ 0:44 live piglets per litter. Aftercross-fostering to balance litter size, sows had an average of 9:4 Æ 0:36 piglets. Weightgain in the first 2 days of life averaged 101 Æ 12:8 g/day per piglet, 239 Æ 6:5 g/day overthe first 20 days and 236 Æ 6:6 g/day to weaning. Sows weaned an average of 8:6 Æ 0:35piglets with a mean weight of 7:8 Æ 0:22 kg at 26 days of age.
In earlier work, we have demonstrated a significant increase in MCR of progester- one with consequent lower plasma progesterone concentrations in response to dou-bling the feed intake of ovariectomised gilts Therefore, we expected to see acomparable reduction in plasma progesterone concentration in high intake pregnantsows between pre-treatment values taken on day 98 of gestation and those on day 109of gestation, particularly in comparison to the concentrations in N sows. Surprisinglythis was not observed. Instead there was a trend towards an increased progesteroneconcentration on day 109 in the H sows. Increased progesterone clearance rate inresponse to increased feeding level is thought to result primarily from increased bloodflow through the liver . If the same mechanism occurred in pregnant sows in thecurrent experiment then there must have been a corresponding increase in productionrate since there was no decline in plasma progesterone concentration. This could occuras the result of negative feedback, although this was not observed in studies with ratsor through increased hormonal stimulation. Both insulin and IGF-1 have beenshown to increase progesterone secretion in a variety of species and both increasewith increasing plane of nutrition. Therefore, high intake feeding of sows in lategestation may result in increased progesterone production as well as increasedclearance. However, it may be that no change in plasma progesterone concentrationwas observed because there was no change in clearance rate in response to increasedfeed intake in pregnant sows.
In the ovariectomized gilt, progesterone concentrations decreased post-prandially in response to increased clearance, presumably via increased splanchnic blood flow inresponse to feed intake. The sows in this experiment were fed only once daily andtherefore feeding might be expected to have a dramatic effect on progesterone clearancerate with a resulting depression in plasma progesterone concentrations measured 2 h afterfeeding on day 109 of gestation when compared to the pre-feeding value. This was notobserved and again there are two possible explanations. Either there was no increasedclearance in response to feeding or secretion of progesterone increased proportionally,perhaps in response to the feeding-induced surge of insulin.
Plasma progesterone concentration in late gestation was not related to resultant litter size but plasma progesterone concentration during farrowing was positively correlated to thisparameter. This suggests that there was some residual progesterone production at farrowingrelated to the litter itself. This is almost certainly due to feto-placental production which will be directly related to fetal number and placental mass. The disappearance rate H.M. Miller et al. / Theriogenology 62 (2004) 1618–1626 constant of progesterone and its half-life in the circulation of sows following farrowingwere of the same order of magnitude as those measured in ovariectomized gilts . Thedisappearance rate constant was positively correlated to plasma progesterone concentra-tion during farrowing, therefore, the lower the plasma progesterone concentration duringfarrowing, the lower the disappearance rate constant and the longer the half-life ofprogesterone.
In this experiment disappearance rate constant was negatively related to sow backfat thickness and therefore fatter sows had longer progesterone half-lives than thinner sows.
The concentration of progesterone in adipose tissue is more than 200 times greater than thatin plasma therefore, following 115 days of elevated plasma progesterone in pregnancy,we would expect progesterone concentrations in fat to be high and similar per unit of fat. Inthis case, fatter sows would have a greater total body content of progesterone than thinnersows because of their greater proportion of fat. Hence, we might expect progesterone to bereleased at a greater rate in fatter sows than in thinner, and possibly over a longer timeperiod, thus sustaining higher levels of progesterone in plasma following farrowing.
Progesterone would therefore have a longer half-life in such animals as was observedin this experiment, surprisingly this was not observed in ovariectomised gilts However, it does appear to make biological sense.
Disappearance rate constant was also negatively correlated to litter size at birth, presumably because sows giving birth to larger litters had greater feto-placental productionof progesterone Hence, plasma progesterone will decline more slowly in sows givingbirth to larger litters.
There was no correlation between the disappearance rate constant of progesterone, or sow plasma progesterone concentration per se, and the onset of lactogenesis as determinedby litter weight gain in the first 2 days of life. This contrasts with the work of de Passille´et al. who observed a positive correlation between plasma progesterone concentrationin sows in the 48 h following farrowing and in 3 days weight gains of piglets. The onlycorrelation between progesterone and piglet weight gain in this experiment was thatbetween the decline in plasma progesterone concentration in the 6 h following farrowingand piglet weight gain to 6 days of age. This does suggest that the more rapidlyprogesterone declines immediately after farrowing, the better the establishment of lacta-tion. However, this effect ceased to be important as the piglets grew to heavier weights.
Increased feed intake in late gestation did not change progesterone concentrations in sows nor increase the rate of decline of progesterone at farrowing. Rate of decline inprogesterone post-farrowing was negatively related to sow fatness. These observations arein direct contrast to findings with ovariectomised gilts and therefore indicate that theprogesterone infused ovariectomised gilt is an inappropriate model in which to studyprogesterone metabolism in the periparturient sow.
H.M. Miller et al. / Theriogenology 62 (2004) 1618–1626 This work was supported by Alberta Agricultural Research Institute (AARI) and Alberta Pork Producers Development Corporation (APPDC). H.M. Miller was supported by aCanadian Commonwealth Scholarship.
The authors wish to thank the personnel of the University Swine Research Centre, especially A. Cegielski, the laboratory and statistical staff in the Department of AFNS.
[1] Ashworth CJ. Effect of pre-mating nutritional status and post-mating progesterone supplementation on embryo survival and conceptus growth in gilts. Anim Reprod Sci 1991;26:311–21.
[2] Hillbrand FW, Elsaesser F. Concentrations of progesterone in the backfat of pigs during the estrous cycle and after ovariectomy. J Reprod Fertil 1983;69:73–80.
[3] Taverne M, Bevers M, Bradshaw JMC, Dieleman SJ, Willemse AH, Porter DG. Plasma concentration of prolactin, progesterone, relaxin and estradiol-17b in sows treated with progesterone, bromocriptine orindomethacin during late pregnancy. J Reprod Fertil 1982;65:85–96.
[4] de Passille´ AMB, Rushen J, Foxcroft GR, Aherne FX, Schaefer A. Performance of young pigs: relationships with periparturient progesterone, prolactin, and insulin of sows. J Anim Sci 1993;71:179–84.
[5] Prime GR, Symonds HW. Influence of plane of nutrition on portal blood flow and the metabolic clearance rate of progesterone in ovariectomized gilts. J Agric Sci 1993;121:389–97.
[6] Miller HM, Foxcroft GR, Squires J, Aherne FX. The effects of feed intake and body fatness on progesterone metabolism in ovariectomized gilts. J Anim Sci 1999;77:3253–61.
[7] Dyck GW, Palmer WM, Simaraks S. Progesterone and luteinising hormone concentration of pregnant gilts on different levels of feed consumption. Can J Anim Sci 1980;60:877–84.
[8] NRC. Nutrient requirements of swine. Washington, DC: National Academy Press; 1988.
[9] Beltranena E, Foxcroft GR, Aherne FX, Kirkwood RN. Endocrinology of nutritional flushing in gilts. Can [10] Ott RL. An introduction to statistical methods and data analysis. Belmont, CA: Duxbury Press; 1992.
[11] Wilkinson L. SYSTAT: the system for statistics. Evanston, IL: SYSTAT Inc.; 1990.
[12] Waddell BJ, Bruce NW, Olynyk JK. Independence of ovarian progesterone secretion rate from arterial progesterone concentrations in the pregnant rat. J Endocrinol 1982;94:61–7.
[13] Braden TD, Belfiore CJ, Niswender GD. Hormonal control of luteal function. In: Findlay JK, editor.
Molecular biology of the female reproductive system. Academic Press, Inc.; 1994. p. 259–87.
[14] Craig VA. Placental steroid metabolism in late pregnancy. In: Cole DJA, Foxcroft GR, editors. Control of pig reproduction. London: Butterworths; 1982. p. 405–18.

Source: http://www.afns.ualberta.ca/en/Research/ProgramsCentres/SRTC/Publications/~/media/afns/Research/SRTC/Documents/Miller_etal_2004_21.pdf