Temporal endocrine and hematological consequences of biological extremes in porcine birth weight

Abstract

To characterize the potential long-term impact of intrauterine growth restriction (IUGR), we assessed the endocrine and hematological consequences of IUGR at birth and in the early postnatal period. Piglets (n = 911) were collected pre-suckling, weighed, and blood sampled, with rectal temperatures assessed two hours later, and sampling repeated at 20 days of age (D20). A subset of piglets was classified as IUGR (n = 92) or large for gestational age (LGA; n = 92) based on birthweight. At birth, IUGR piglets had decreased T3 and T4 relative to LGA, which may be associated with concomitantly decreased rectal temperatures. T4 had normalized by D20, while T3 was increased in IUGR relative to LGA at this timepoint. Cortisol did not differ between groups at either timepoint. IUGR piglets also experienced decreased hematocrit, blood glucose, and plasma protein at birth, while only plasma protein remained decreased at D20. At birth, male IUGR piglets had decreased testosterone compared to LGA piglets, while estradiol levels were not altered in female IUGR piglets. While IUGR piglets exhibited catch-up growth, they failed to reach an appropriate weight by D20. These results suggest that some of the systemic dysregulations occurring in IUGR piglets have normalized by D20, even though bodyweight fails to recover.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-27620-7.

Introduction

Intrauterine growth restriction (IUGR) is a common developmental disorder originating in fetal life and leading to the birth of small, unthrifty infants. While the term IUGR is often used synonymously with small for gestational age (SGA), the majority of cases of IUGR result in the birth of an infant that is not only small, but also experiences non-allometric growth, with growth restriction of organs such as the liver leading to an abnormally reduced abdominal circumference¹. IUGR may result from various maternal or fetal factors, with the most common cause being insufficient or abnormal placentation¹. Regardless of origin, IUGR offspring are known to experience a wide array of physiological consequences both at birth and in adulthood, including impaired organ development^2,3, an increased propensity to develop disease⁴, and endocrine^5,6 and hematologic dysregulations⁷.

To aid in studying the consequences of IUGR, many biomedical models have been developed in a variety of species, often involving nutritional⁸, thermal^9,10, or surgical^11,12 manipulation of the dam during the gestational period. In addition to these artificially induced models of IUGR, IUGR is known to occur spontaneously in contemporary swine as a result of selection for litter size well in excess of uterine capacity, which causes fetal crowding and placental insufficiency¹³. Prior studies report the incidence of naturally occurring IUGR in swine to be around 20–30%^14,15, with many studies utilizing low birth weight as an identifier of porcine IUGR^16,17. For this reason, swine are a valuable biomedical model for studying the impact of naturally occurring IUGR in the absence of experimental manipulation, with low birth weight piglets displaying the characteristic non-allometric body and organ growth of asymmetrically IUGR human infants¹⁸.

Many systemic endocrine and hematological consequences of IUGR have previously been noted in IUGR human infants and piglets. For example, both human and porcine IUGR offspring are known to have altered levels of insulin-like growth factor 1^5,18, insulin^19,20, and thyroid hormones^6,16. Further work in human infants has shown altered levels of angiotensin II²¹, a vasoconstrictor, as well as altered levels of the major hormones involved in stimulating release of the stress hormone cortisol²². Additionally, both human and porcine cases of IUGR are characterized by a multitude of hematologic dysregulations in the early postnatal period^7,20,23. While these endocrine and hematological consequences are largely understood at birth, a longer-term study examining whether or not these alterations persist past the neonatal period is warranted, considering the known impact of IUGR on health into adulthood⁴. Consequently, the pig is a beneficial biomedical model as it allows for a longitudinal study under more controlled environmental conditions than could be maintained in humans.

To this end, the objective of the present study was to examine the temporal endocrine and hematological consequences of IUGR in a spontaneously occurring porcine biomedical model. Our hematologic targets included hematocrit, blood glucose, and plasma protein, while our endocrine targets included cortisol and the two major thyroid hormones (T3 and T4), the latter of which are critical regulators of metabolism, growth, and development. Additionally, we assessed neonatal levels of testosterone (T) and 17β-estradiol (E2) in male and female IUGR piglets, respectively, as prior research has suggested a link between low birth weight, IUGR, and impaired reproductive development^24–28, but to the author’s best knowledge, has yet to examine sex steroid concentrations in the immediate neonatal period in the pig. We hypothesized that in comparison to piglets born large for gestational age (LGA), IUGR piglets would exhibit endocrine and hematologic dysregulations at birth, with these dysregulations persisting throughout the lactational period. To assess this, litters of piglets were weighed and blood sampled at birth and 20 days of age. Birth weight was subsequently used to assign a subset of piglets into IUGR and LGA groups, and the temporal impact of these classifications on the various endocrine and hematologic parameters assessed.

Materials and methods

Animal model and sample collection

A total of n = 83 pregnant commercial sows between the first and eighth parity (2.0±1.0 (mean±SD)) were selected from and housed at the Purdue University Animal Sciences Research and Education Center (ASREC). Prior to this selection, the n = 83 animals had been bred in three sequential production batches via artificial insemination using pooled Duroc semen, and pregnancy confirmed by transcutaneous ultrasound between gestational days 25 and 35 in accordance with farm standard operating procedures. Sows were moved into farrowing crates on gestational day 110.8±0.9, and continually monitored at 15 min intervals in the days prior to expected farrowing (monitoring beginning at gestational day 112.3±0.8). Sows were allowed to farrow naturally until gestational day 116, at which point they were induced with an injection of 10 mg dinoprost (LUTALYSE; Zoetis, Parsippany, NJ, USA), followed by 10 USP units of oxytocin (Bimeda, Schaumburg, IL, USA) administered 6 and 12 h later. Thirty-one sows were induced, with the average gestational length across all three batches being 115.2±1.3. While the farm this study was conducted on is not biosecure by industry standards and has previously observed cases of PRRSV1, PCV2, PCV3, and mycoplasma, all sows appeared healthy throughout the course of the experiment.

A total of 1025 (498 males and 527 females) piglets were successfully collected prior to suckling, with any piglet that may have suckled excluded from the study. Piglets were immediately towel-dried, weighed, sexed, and ear tagged. Piglets were placed in dorsal recumbency and approximately 3 mL of EDTA plasma was successfully collected from n = 911 (453 males and 458 females) piglets by venipuncture of the external jugular vein, after which the piglets were immediately returned to the sow’s udder, which had a standard heat lamp hung over top of it. Two hours later, piglet rectal temperatures were taken and recorded. The average total born litter size across all three batches was 14.4±3.6, with the average number of liveborn piglets per litter being 13.4±3.7. Following farrowing, litter sizes were normalized within each batch via cross-fostering. Piglets were managed throughout the lactational period according to the farm’s standard operating procedures, with ear-notching, tail-docking, teeth-clipping, injection of 160 mg of iron (Gelptoforte^®, Ceva Animal Health, Lenexa, Kansas, USA), and castration of healthy males occurring in the first 24 h of life. Additionally, piglets were treated and/or euthanized as needed according to farm staff’s discretion, with euthanasia accomplished by carbon dioxide inhalation. At 20 days of age (D20; 19.5±1.3), the remaining n = 806 piglets were reweighed and n = 754 successfully blood-sampled according to the procedure originally conducted at birth. D20 was selected to avoid the potential physiological impact of the stress associated with weaning, which was conducted on D21 after termination of the trial. Following D20 sampling, average daily gain (ADG) was calculated as the average weight gained per day throughout the lactational period (g/day), with fractional ADG calculated by dividing the ADG by the piglet’s initial birth weight (g/kg/day). Following whole blood measurements, blood was centrifuged at 783 × g for 15 min, and the resultant plasma removed and stored at -20 °C. All animal procedures are reported in accordance with the ARRIVE guidelines and were conducted in compliance with Purdue University’s animal care policies, approved by the Institutional Animal Care and Use Committee (Protocol #2103002122).

Subset selection

To assess the consequences of IUGR, a subset of biological extremes (n = 184) was selected for analysis based on the top and bottom 10th percentiles of birth weight. Piglets were first split based on sex, with the lowest birth weight males (n = 46) and females (n = 46) selected to represent IUGR piglets. A contrasting population of the highest birth weight males (n = 46) and females (n = 46) was selected to represent large for gestational age (LGA) piglets, which are less likely to have experienced in utero crowding than piglets of an average birth weight. All unselected piglets were subsequently labeled as appropriate for gestational age (AGA). Following selection, one LGA female was removed from the analyses due to abnormally poor performance during the lactational period, as noted by marked weight loss from birth to D20. At D20, the populations for analysis were reduced, largely due to piglet death throughout lactation. The final sample sizes measured for all assays at birth and D20 are represented in Table 1, with the sample sizes for all weight data represented in Table 2. Prior to data analysis, hematological data from one female LGA piglet at D20 was removed from the data set, as the blood was recorded as clotted.

Table 1.

Sources, product identifiers, coefficients of variance (CVs), and sample sizes for study endocrine and hematological assays.

Endocrine and Hematological Assays					Birth Sample Size (M/F)1		D20 Sample Size (M/F)1
Assay	Source	Product Identifier	Intra-assay CV	Inter-assay CV	IUGR	LGA	IUGR	LGA
Triiodothyronine (T3)	MP Biomedicals	07M175-CF	3.76	8.49	92(46/46)	91(46/45)	53(24/29)	82(41/41)
Thyroxine (T4)	MP Biomedicals	07M275-CF	3.25	7.07	92(46/46)	91(46/45)	53(24/29)	82(41/41)
Cortisol	MP Biomedicals	07M3675A	4.88	14.35	92(46/46)	91(46/45)	53(24/29)	82(41/41)
Testosterone (T)	MP Biomedicals	07M3775A	6.45	9.49	46(46/0)	46(46/0)	NA	NA
17β-Estradiol (E2)	MP Biomedicals	07M4975A	4.09	8.63	46(0/46)	45(0/45)	NA	NA
Protein	Thermo Fisher	23,225	3.07	4.27	92(46/46)	91(46/45)	53(24/29)	82(41/41)
Glucose	Germaine Laboratories	37,321	–	–	90(45/45)	91(46/45)	54(24/30)	81(40/41)
Hematocrit	–	–	–	–	89(44/45)	90(46/44)	53(24/29)	84(42/42)

¹Sample sizes for intrauterine growth restricted (IUGR) and large for gestational age (LGA) groups listed as overall value first, with breakdown of males (M) and females (F), respectively, shown in parentheses.

Table 2.

Phenotypes, rectal temperatures, and growth rates of male and female piglets divided into intrauterine growth restricted (IUGR), appropriate for gestational age (AGA), and large for gestational age (LGA) groups based on birth weight.

Group	IUGR		AGA		LGA		P-Values
Sex	Male	Female	Male	Female	Male	Female	Group	Sex
Birth Sample Size (n)	46	46	361	366	46	45	NA	NA
D20 Sample Size (n)	26	36	322	335	42	44	NA	NA
Birth Weight (kg)	0.758 ± 0.100a	0.767 ± 0.131a	1.338 ± 0.234b	1.310 ± 0.206b	1.924 ± 0.133c	1.854 ± 0.179c	< 0.001	0.032
D20 Weight (kg)	3.957 ± 0.783a	3.948 ± 0.747a	5.554 ± 1.005b	5.513 ± 0.997b	6.272 ± 1.229c	6.360 ± 1.081c	< 0.001	0.724
2 h Rectal Temperature (℃)	36.617 ± 1.826a	36.995 ± 1.757a	38.201 ± 0.788b	38.217 ± 0.846b	38.570 ± 0.854c	38.492 ± 0.666c	< 0.001	0.509
ADG (g/day)1	164.626 ± 39.590a	159.588 ± 37.124a	216.442 ± 47.595b	215.694 ± 46.697b	228.462 ± 60.927c	237.836 ± 51.790c	< 0.001	0.997
Fractional ADG (g/kg/day)2	207.742 ± 49.607a	203.231 ± 42.470a	164.621 ± 40.113b	167.550 ± 39.436b	118.850 ± 31.835c	128.784 ± 28.679c	< 0.001	0.262

¹Average daily gain (ADG) was calculated as the average daily weight gain between birth and D20 sampling, and²fractional ADG calculated as the ADG divided by the birth weight. Data was analyzed as a two-way ANOVA including group and sex, with pairwise group comparisons made using Tukey’s HSD test. ^{a, b,c}Significant differences (P < 0.05) between the IUGR, AGA, and LGA groups are denoted by unique letter superscripts. Data are presented as mean ± SD.

Endocrine assays

Endocrine assays for circulating 3,5,3’-triiodothyronine (T3), thyroxine (T4), and cortisol were conducted on previously selected subsets at birth and D20 using commercial chemiluminescent immunoassay kits according to manufacturer’s directions (MP Biomedicals, Solon, OH, USA). Briefly, 50 (T3) or 25 (T4 and cortisol) µL of plasma was assayed in duplicate and the luminescence quantified using a Spark 10 M spectrophotometer (Tecan Life Sciences, Männedorf, Switzerland). Endocrine assays for circulating T and E2 were also conducted using commercially available chemiluminescent immunoassay kits (MP Biomedicals), with T measured only on male samples at birth, and E2 measured only on female samples at birth. As before, all samples were assayed in duplicate and the luminescence quantified using a spectrophotometer, however, both assays were performed with minor modifications to the manufacturer’s protocols. Namely, to effectively assess the low plasma T concentrations seen in neonatal piglets, sample volume for the T assay was tripled. For measurement of E2, 25 µL of plasma sample was assayed as directed, but results were analyzed relative to an abbreviated 5-point standard curve (73.42-5,506.5 pmol/L). For all hormones, each assay included two independent sample pools run on each plate which were used to calculate the inter-assay variance, with the average coefficients of variance for each assay shown in Table 1.

Hematological assays

Measurements of hematocrit, blood glucose, and plasma protein were also performed at both birth and D20. Hematocrit and blood glucose were conducted on whole blood samples from all piglets immediately following collection, with only the subset of data representing IUGR and LGA piglets analyzed. The assays for both hematocrit and glucose were shown to be reproducible over repeated measurements in a preliminary experiment (Supplementary Table S1), and as a result, these parameters were measured only in singlet. For hematocrit, 60 µL of whole blood was pipetted into a microhematocrit tube and one end sealed with wax. The microhematocrit tube was then centrifuged at 1534 × g for 5 min and read on a standard reader card. For measurement of glucose, a droplet of whole blood was assessed using AimStrip Plus blood glucose strips and meters (Germaine Laboratories, Inc., San Antonio, TX, USA). Plasma protein was assessed on IUGR and LGA piglets using the Pierce BCA Protein Assay Kit according to manufacturer’s directions (Thermo Fisher, Waltham, MA, USA). Briefly, plasma samples were diluted 1:100, and 25 µL of each sample assayed in duplicate. The plate was incubated at 37 °C for 30 min, and the resultant absorbance measured on a Spark 10 M spectrophotometer (Tecan Life Sciences). Two independent sample pools were run on each plate and used to calculate the inter-assay variance, with the average inter- and intra-assay coefficients of variance shown in Table 1.

Statistical analyses

All statistical analyses were performed in R version 4.3.1²⁹, with data visualization accomplished using the ggplot2 package³⁰. Probability density functions were calculated using the dnorm function in base R. Data for T and E2 were assessed using a type III one-way ANOVA, while all other data were assessed using a type III two-way ANOVA including group and sex. The interaction between group and sex was initially included in all models and found to be insignificant and removed for all parameters except for blood glucose at D20. For the rectal temperature data, four IUGR piglets (two males and two females) had temperatures that were below the 32 °C lower limit of the rectal thermometer despite two repeated attempts at measurement, so were assigned a temperature of 32 °C for analysis. For all data, normality was first visually assessed and data log-transformed to improve normality as needed, with all results presented on the original scale. For assessments reported in Table 2, appropriate post-hoc pairwise comparisons were made using Tukey’s HSD test. For reporting of assay data, the absolute difference between the LGA and IUGR median values was calculated and abbreviated as Δx̃. The threshold for statistical significance was P < 0.05.

Results

Piglet Phenotypes, rectal temperatures, and growth rates

The probability density functions for piglet weights at birth and D20 are shown in Fig. 1. Visual assessment of the distribution at birth displays the population average birth weight of 1.323 kg, with piglets selected as IUGR spanning a birth weight of 0.392 kg to 0.921 kg (x̅=0.762 kg), and piglets selected as LGA spanning a birth weight of 1.679 kg to 2.392 kg (x̅=1.889 kg) (Fig. 1A). Assessment of the distribution at D20 shows a mean population weight of 5.495 kg, with a regression towards the mean observed at varying degrees for both the IUGR (x̅=3.952 kg, 1.85–5.35 kg) and LGA (x̅=6.317 kg, 3.30–9.20 kg) groups, with some LGA piglets falling below the average weight, but no IUGR piglets growing above the average weight (Fig. 1B).

Fig. 1.

Piglet body weight distributions. Probability density functions of piglet weights at (A) birth (n = 910) and (B) 20 days of age (n = 805). Point shape and color represent group and sex, with red triangles representing piglets selected as intrauterine growth restricted (IUGR; below the ~ 10th percentile of birth weight within each sex), green squares representing piglets selected as large for gestational age (LGA; above the ~ 90th percentile of birth weight within each sex), and greyscale circles representing appropriate for gestational age piglets (AGA; between the 10th and 90th percentile of birth weight within each sex).

A statistical assessment of birth weight, D20 weight, rectal temperature, ADG, and fractional ADG between IUGR, AGA, and LGA piglets is shown in Table 2. Consistent with selection criteria, both birth weight and D20 weight were significantly different between all three groups, being largest in the LGA group and smallest in the IUGR group (P < 0.001 for all comparisons). In addition, male piglets were found to be significantly heavier at birth compared to female piglets, regardless of group, with a marginal difference between means of 0.029 kg (P = 0.032). Like body weight, the rectal temperature value recorded at 2 h after placement at the sow was also significantly different between all three groups, with rectal temperature highest in LGA piglets and lowest in IUGR piglets (P < 0.001 for IUGR vs. AGA and IUGR vs. LGA; P = 0.007 for AGA vs. LGA). Additionally, ADG and fractional ADG were significantly different between all three groups. ADG was highest in LGA piglets and lowest in IUGR piglets (P < 0.001 for IUGR vs. AGA and IUGR vs. LGA; P = 0.005 for AGA vs. LGA). In contrast, fractional ADG was highest in IUGR piglets and lowest in LGA piglets (P < 0.001 for all comparisons), suggesting compensatory postnatal growth.

Endocrine profiles at birth and D20

At birth, concentrations of T3 and T4 were significantly reduced in IUGR piglets relative to LGA (Fig. 2A and B), with a Δx̃ of 0.111 nmol/L for T3 (P = 0.029), and a Δx̃ of 21.477 nmol/L between groups for T4 (P < 0.001). In contrast, there was no significant difference in cortisol levels observed between LGA and IUGR piglets (Fig. 2C) at birth. Opposite of the findings at birth, IUGR piglets at D20 had increased circulating T3 levels relative to LGA piglets, with a Δx̃ of 0.281 nmol/L (P = 0.009) (Fig. 2D). The other endocrine parameters remeasured at D20, T4 and cortisol, did not vary between IUGR and LGA piglets (Fig. 2E and F) at this timepoint. Of all endocrine measurements at birth and D20, T4 levels at D20 were the only measurement significantly impacted by sex regardless of group, with D20 T4 levels significantly higher in female piglets relative to male piglets (Δx̃=4.440 nmol/L; P = 0.008).

Fig. 2.

Endocrine profiles of LGA and IUGR piglets at birth and 20 days of age. Circulating hormone levels of (A, D) triiodothyronine (T3), (B, E) thyroxine (T4), and (C, F) cortisol in LGA and IUGR piglets at (A, B, C) birth (n = 183) and (D, E, F) 20 days of age (D20; n = 135). Male and female piglets are represented as orange and purple points, respectively. Data was analyzed as a two-way ANOVA including group and sex, with statistical differences between the LGA and IUGR groups denoted as *p < 0.05, **p < 0.01, and ***p < 0.001.

Hematologic profiles at birth and D20

In comparison to LGA piglets, IUGR piglets also had decreased hematocrit, blood glucose, and plasma protein at birth. This difference was most severe for blood glucose, with a Δx̃ of 0.916 mmol/L (P < 0.001) (Fig. 3B). In contrast, the Δx̃ for hematocrit was only 2% (P < 0.001) (Fig. 3A), while plasma protein was marginally decreased by a Δx̃ of only 1.468 g/L (P = 0.011) (Fig. 3C). At D20, plasma protein remained decreased in IUGR piglets relative to LGA piglets, with a Δx̃ of 3.889 g/L (P = 0.020) (Fig. 3F). Conversely, the reduced hematocrit and blood glucose levels observed in IUGR piglets at birth appeared resolved at D20 (Fig. 3D and E), with no differences observed between groups at this timepoint. Of all the hematological parameters assessed at birth and D20, sex, regardless of group, only had a significant impact on hematocrit at birth, with female piglets having a significantly higher hematocrit percentage than male piglets (Δx̃=2%; P = 0.017). Additionally, a significant interaction between group and sex was observed for blood glucose levels at D20 only (P = 0.033), with a significant increase in blood glucose levels observed in female IUGR piglets relative to female LGA piglets (Δx̃=0.416 mmol/L; P = 0.048), and no difference observed between the equivalent male groups.

Fig. 3.

Hematologic profiles of LGA and IUGR piglets at birth and 20 days of age. Measurements of (A, D) hematocrit, (B, E) blood glucose, and (C, F) plasma protein in LGA and IUGR piglets at (A, B, C) birth (n = 179–183) and (D, E, F) 20 days of age (D20; n = 134–136). Male and female piglets are represented as orange and purple points, respectively. Data was analyzed as a two-way ANOVA including group and sex, with statistical differences between the LGA and IUGR groups denoted as *p < 0.05, **p < 0.01, and ***p < 0.001.

Sex steroid concentrations at birth

At birth, male IUGR piglets had lower circulating T than male LGA piglets (Fig. 4A), with a Δx̃ of 0.465 nmol/L (P < 0.001). In contrast, there was no difference in E2 concentrations observed between female IUGR and female LGA piglets (Fig. 4B).

Fig. 4.

Sex steroid concentrations in LGA and IUGR piglets. (A) Levels of circulating testosterone (T) in male LGA and IUGR piglets at birth (n = 92), and (B) levels of circulating 17β-estradiol (E2) in female LGA and IUGR piglets at birth (n = 91). Data was analyzed as a one-way ANOVA, with statistical differences between the LGA and IUGR groups denoted as *p < 0.05, **p < 0.01, and ***p < 0.001.

Discussion

The present study utilized a porcine model of spontaneously occurring IUGR to investigate the impact of this developmental perturbation throughout the early postnatal period. In humans, IUGR can manifest either symmetrically, whereby the whole body is proportionally reduced in size, or asymmetrically, whereby non-allometric growth is observed. However, asymmetric IUGR is the most common type and accounts for around 70–80% of human cases¹. In swine IUGR is typically asymmetric¹⁸, characterized by altered cranial features and high brain to liver weight ratio. The condition in swine is a consequence of selective breeding for increased litter sizes¹³, which has increased ovulation rate beyond uterine capacity resulting in placental insufficiency. Consistent with many prior studies, the present study found that IUGR is associated with a wide array of endocrine and hematological consequences, however, we further report that some of these effects have normalized by D20 when assessed in the same piglet population and compared to piglets born LGA. This suggests that some of the systemic effects of IUGR on endocrinology and hematology relative to LGA piglets do not persist long-term.

Perhaps the most novel finding of this study is that IUGR piglets have reduced T3 and T4 at birth in comparison to their LGA counterparts. This reduction in T3 and T4 mimics what has previously been observed in human fetuses with IUGR⁶, but contradicts one older report in the neonatal piglet which showed that IUGR piglets have increased circulating T4 levels relative to normal birth weight piglets¹⁶. In addition to utilizing a much smaller sample size, this prior study utilized a sampling procedure involving anesthesia with isoflurane¹⁶, which has previously been shown to impact thyroid hormone levels in humans³¹. Alternatively, it is also possible that the discrepancy in results may be attributed to the choice of comparison group, as our study compared IUGR piglets to LGA piglets, which exhibited higher rectal temperatures and ADG than AGA piglets, suggesting potential metabolic differences between these two groups. Regardless, one possible explanation for a reduction in T3 and T4 levels could be a developmental delay in thyroid function, as it is well documented that, throughout gestation, ontogenic increases in concentration of both major thyroid hormones occur in the human⁶ and the pig fetus³². As such, the reduction in T3 and T4 observed in IUGR piglets in the present study may be attributed to delayed maturation of the thyroid gland relative to LGA piglets.

In addition to reduced T3 and T4, IUGR piglets in the current study had reduced rectal temperatures at two hours postnatally, indicating an impaired thermoregulatory capability. This is consistent with prior reports in the piglet³³, with an increased risk of hypothermia also noted in human infants with IUGR¹. Consistent with the rest of our data, data from prior reports suggests a potential relationship between low glucose levels and impaired thermoregulation in IUGR³⁴. As thyroid hormones play a major role in regulating basal metabolic rate, another possible, or concomitant, explanation for impaired thermoregulation in IUGR neonates may be the observed deficit in T3 and T4. Prior attempts to increase rectal temperatures in IUGR piglets by administering glucose injections have proven ineffective³⁵, with one possible limitation being the short half-life of circulating glucose. While not previously examined, one could hypothesize that T4 supplementation to IUGR neonates may be a better alternative, as T4 supplementation is known to increase body temperature in mice³⁶. Additionally, T4 has a much longer half-life than glucose, meaning a single dose given postnatally may be enough to increase body temperature in the critical neonatal period prior to the normalization of thyroid hormone levels and body temperature later in life.

The finding of reduced hematocrit in IUGR piglets relative to LGA is consistent with prior reports showing decreased hematocrit in IUGR piglets relative to normal birth weight piglets at 3 days of age²⁰. However, these findings contrast what is observed in the majority of human infants with IUGR, who exhibit increased hematocrit and polycythemia¹. One potential explanation for this species discrepancy may be the varying placental structures and mechanisms of iron transfer. Piglets are born with very low iron levels, as the transfer of maternal iron across the diffuse epitheliochorial placenta is spatially restricted to the areolae³⁷. Total placental areolae number is known to be decreased in the placenta of small fetuses¹², suggesting that the reduced hematocrit seen in IUGR piglets at birth may result from an exacerbated iron deficit beginning in utero. In contrast to the pig, the hemochorial placenta of the human allows for a much more robust and efficient transfer of iron. Consequently, the human infant is born with an iron store of around 250 mg³⁸, while the piglet is born with iron reserves less than 50 mg³⁹. IUGR is known to lead to chronic fetal hypoxia, which increases erythropoietin levels in the human⁴⁰. In the presence of sufficient iron, as is the case in the human fetus, erythropoietin can drive red blood cell production, leading to polycythemia. Due to the deficiency of iron in the porcine fetus and neonate, an equivalent increase in erythropoietin may fail to cause any marked red blood cell production, effectively explaining this discrepancy between the human and the pig. However, the routine administration of an iron injection to neonatal piglets, as was utilized in the present study, may contribute to the normalization of hematocrit by D20, making the blood profile of IUGR piglets more reminiscent to that of the human.

The present study observed decreased total plasma protein in IUGR piglets at both birth and D20, making this measure the only parameter exhibiting a sustained decrease in IUGR piglets relative to LGA. Additionally, the data reiterates prior reports that piglets are born relatively hypoproteinemic⁴¹, as the mean protein concentration measured at birth was nearly half of that measured at D20. One potential explanation for the decrease in total protein in IUGR piglets relative to LGA piglets could be the reduction in liver size observed in IUGR, which is known to persist throughout the lactational period in the pig⁴². As the liver is the primary producer of the majority of blood proteins, one could surmise that a decrease in overall liver size would decrease production of key proteins such as albumin, leading to the observed decrease in overall plasma protein. Our results are consistent with a prior study that reported long-term decreases in total serum protein in IUGR piglets relative to normal birth weight piglets, with this finding marked by concomitant increases in serum aminotransferase concentrations, indicating potential liver dysfunction and altered protein metabolism⁴³. Offspring with IUGR are additionally known to have altered immune function⁴⁴, which could be partially attributed to a reduction in circulating immunoglobulin content. This reduction has previously been observed in human infants with IUGR⁴⁵, and could also be a contributing factor to the overall reduction in total plasma protein observed at D20 in the present study. However, this would not explain the low plasma protein levels observed prior to suckling, as the restrictive epitheliochorial placenta of the pig prevents transfer of immunoglobulins from dam to fetus⁴⁶, meaning piglets are born agammaglobulinemic. In addition, while not assessed in the present study, a previous study conducted among IUGR and normal birth weight piglets reported no significant differences in plasma concentrations of IgG, IgA, or IgM when measured at 3 days of age²⁰.

Interestingly, the present study also observed that male IUGR piglets had reduced T levels at birth relative to LGA, while E2 levels in female IUGR piglets were unaffected. The sensitivity of male reproductive development to the effects of growth restriction has previously been demonstrated in both humans and pigs. Humans born small have decreased post-pubertal testicular weight and circulating T levels²⁵, as well as an increased risk of infertility⁴⁷, while low birth weight pigs have decreased testicular weight, decreased daily sperm production, and altered seminal plasma composition in adulthood^27,28. Interestingly, prior reports in the pig indicate that circulating T is not decreased in male low birth weight piglets when measured at either 8 days or 8 months of age and compared to T levels in high birth weight piglets²⁸. One potential explanation for this discrepancy is that the low birth weight piglets selected in the previous study were nearly 250 g larger at birth than our current selection of IUGR piglets. Alternatively, the discrepancy in results may suggest low T in the IUGR pig is temporally specific and restricted to the in utero and immediate postnatal period. Despite the potentially acute nature of this decrease in T, the sensitivity of the developing fetus to endocrine perturbation may still allow for long-term reproductive impacts. In the porcine testes, rapid hyperplasia of the Sertoli and Leydig cells is known to occur in the first month postnatally, with volume growth of the Leydig cells following a similar trajectory to circulating T levels⁴⁸. In the absence of sufficient T in this critical neonatal period, one could hypothesize that testicular development may be altered, with one prior study showing that porcine IUGR results in reductions in testicular volume, distal testis Sertoli cell counts, and the number of Leydig cells²⁷.

IUGR has often been associated with fetal programming leading to long-term health detriments and an increased propensity to develop diseases in adulthood⁴. In the present study, there was evidence of attempted compensatory growth in the IUGR group, as indicated by increased fractional ADG. However, IUGR piglets failed to reach an appropriate body weight by D20, meaning they remained significantly lighter than AGA piglets of the same age. This attempt at catch-up growth is consistent with what is observed in IUGR human infants, and is associated with metabolic programming that increases the risk for developing metabolic diseases later in life⁴⁹. While most endocrine and hematological parameters measured in the current study had normalized in IUGR piglets relative to LGA by D20, there was a unique increase in T3 observed in IUGR piglets relative to LGA at this timepoint. As the majority of thyroid hormone produced directly by the thyroid gland is T4, an alteration in T3 alone is likely to indicate differences in peripheral metabolism, rather than central regulation. As such, the rebound of most endocrine and hematological measurements by D20 suggests that the central regulation of these parameters is not programmed by in utero growth restriction.

One potential limitation of this study is the method utilized for IUGR identification. The current gold-standard for identifying IUGR in swine is the brain-to-liver weight ratio, which is increased in IUGR piglets due to the established brain-sparing effect. As this metric cannot be directly assessed in a live animal, other identification methods including birth weight or subjective scoring of phenotypic features are commonly utilized (reviewed by Ruggeri et al.⁵⁰). While scoring methods based on head shape may better differentiate between piglets that are born IUGR versus SGA, the subjective nature of this approach raises concerns about its reproducibility. For example, two prior studies utilizing the same scoring method for IUGR identification across pigs of the same breed report a more than 10% discrepancy in the percentage of piglets being classified as IUGR^14,15. As such, the need to develop quantitative and objective morphometrics for IUGR identification, such as those recently described by Jeon et al.⁵¹, remains. In light of this, the present study chose to classify piglets as IUGR based on birth weight alone, although it should be recognized that in addition to true IUGR piglets, this group may also include a subset of piglets that are only SGA (meaning they are constitutively small but do not have IUGR). While a distinction between true IUGR and SGA cannot be made in the present study, it is important to recognize that IUGR is more closely associated with adverse postnatal outcomes, as indicated by a variety of reports in the human^52,53.

Another important limitation of this study is the choice of LGA piglets, rather than AGA piglets, as the comparison group for endocrine and hematological measures. As IUGR is not binary and thus occurs along a gradient of varying severities, the LGA group was chosen based on the rationale that LGA piglets are less likely than AGA piglets to have experienced any appreciable degree of growth restriction. This allowed for a truer comparison of the effects of IUGR than would be achieved by comparing IUGR piglets to AGA piglets. However, like in IUGR, data in humans shows that individuals born LGA may also experience physiological deviations from normality⁵⁴, with LGA piglets in the present study exhibiting significantly increased neonatal rectal temperatures and pre-weaning ADG, and a significantly decreased fractional ADG, relative to AGA piglets.

Conclusions

In conclusion, the results of the present study demonstrate that neonatal IUGR piglets experience endocrine dysregulations relative to LGA piglets, including reduced circulating T3 and T4, which may contribute to impaired thermoregulation and is consistent with what is observed in human infants with IUGR. IUGR in the neonatal piglet is further associated with reduced male testosterone levels, as well as hematologic dysregulations including reduced hematocrit, blood glucose, and plasma protein levels. At D20, the endocrine and hematologic profiles of IUGR piglets have largely recovered relative to LGA piglets, suggesting that some systemic consequences of IUGR are acute and do not persist into the early postnatal period, despite IUGR being associated with long-term adverse health impacts.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

JAP and JSR conceived of the study and secured funding. JAP, JSR, and AAS established the initial hypotheses. The animal trial was conducted by AAS, DJ, KR, SI, and JAP. AAS conducted the laboratory work, statistical analyses, and drafted the manuscript, which was reviewed and approved by all authors.

Funding

This study was supported by the intramural research program of the U.S. Department of Agriculture, National Institute of Food and Agriculture, Agriculture and Food Research Initiative [2023-67015-39338]. The findings and conclusions have not been formally disseminated by the U.S. Department of Agriculture and should not be construed to represent any agency determination or policy.

Data availability

The data supporting the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.