Fetal body weight and the development of the control of the cardiovascular system in fetal sheep

Authors


Corresponding author M. Schwab: Department of Neurology, Friedrich Schiller University, D-07740 Jena, Germany. Email: matthias.schwab@med.uni-jena.de

Abstract

Reduced birth weight predisposes to cardiovascular diseases in later life. We examined in fetal sheep at 0.76 (n= 18) and 0.87 (n= 17) gestation whether spontaneously occurring variations in fetal weight affect maturation of autonomic control of cardiovascular function. Fetal weights at both gestational ages were grouped statistically in low (LW) and normal weights (NW) (P < 0.01). LW fetuses were within the normal weight span showing minor growth dysproportionality at 0.76 gestation favouring heart and brain, with a primary growth of carcass between 0.76 and 0.87 gestation (P < 0.05). While twins largely contributed to LW fetuses, weight differences between singletons and twins were absent at 0.76 and modest at 0.87 gestation, underscoring the fact that twins belong to normality in fetal sheep not constituting a major malnutritive condition. Mean fetal blood pressure (FBP) of all fetuses was negatively correlated to fetal weight at 0.76 but not 0.87 gestation (P < 0.05). At this age, FBP and baroreceptor reflex sensitivity were increased in LW fetuses (P < 0.05), suggesting increased sympathetic activity and immaturity of circulatory control. Development of vagal modulation of fetal heart rate depended on fetal weight (P < 0.01). These functional associations were largely independent of twin pregnancies. We conclude, low fetal weight within the normal weight span is accompanied by a different trajectory of development of sympathetic blood pressure and vagal heart rate control. This may contribute to the development of elevated blood pressure in later life. Examination of the underlying mechanisms and consequences may contribute to the understanding of programming of cardiovascular diseases.

Numerous epidemiological and experimental studies demonstrate an association between reduced birth weight and increased likelihood of elevated blood pressure and cardiovascular diseases in later life (for reviews see Barker, 1998; Huxley et al. 2000; Poston, 2001; Langley-Evans et al. 2003; Schwartz & Thornburg, 2003; Adair & Dahly, 2005; McMillen & Robinson, 2005). This association is explained by the fetal programming hypothesis proposing that adverse environmental influences during critical stages of fetal development such as insufficient nutrient supply lead to fetal adaptations that result in both reduced fetal growth and permanent changes in the activity of functional systems. Examples of such changes are a resetting of the set-point of the hypothalamo-pituitary-adrenal (HPA) axis, with enhanced stress responses during later life, adaptation of the metabolic system with decreased capacity and increased efficiency, altered function of the insulin-like growth factor, prostaglandin and renin–angiotensin system, with altered nephrogenesis, and complex alteration of vasoactive endothelium-dependent and -independent mediator systems resulting in increased vascular tone (Phillips et al. 2001; Edwards et al. 2001; McMillen & Robinson, 2005). Programming of altered autonomic control of cardiac and cardiovascular function has been demonstrated in individuals who were small for gestational age (Phillips et al. 2001; Ijzerman et al. 2003). Changes in the autonomic control are a conditio sine qua non for a permanent blood pressure increase and may be due to an altered development of autonomic control mechanisms and not only an adaptation of the autonomic nervous system during later life. Different trajectories of the development of autonomic control mechanisms should be detectable as soon as these mechanisms mature.

In sheep, a widely used model of human fetal development, the last third of gestation is the critical time window of autonomic (Nuwayhid et al. 1975; Walker et al. 1978) and cardiovascular (Blanco et al. 1988; Crowe et al. 1995; Unno et al. 1999; Shinozuka et al. 2000) maturation (Crowe et al. 1995; Unno et al. 1999; Shinozuka et al. 2000). There are only a few studies with contradictory results on the relationship of fetal growth, elevated blood pressure and its autonomic control in prenatally developing species (Cock & Harding, 1997; Murotsuki et al. 1997; Hawkins et al. 2000a; Hawkins et al. 2000b; Louey et al. 2000; Edwards & McMillen, 2001; Edwards et al. 2001). Similar contradictory results are observed in offspring after prenatal nutrient restriction (Hawkins et al. 2000a; Louey et al. 2000; Oliver et al. 2002; Gopalakrishnan et al. 2004; Gilbert et al. 2005). Studies on the association of nutrition and development of autonomic blood pressure control are even rarer. Reduced fetal weight (Crowe et al. 1995) or malnutrition in early gestation without inducing reduced fetal weight is associated with altered baro- and chemoreflex responses in late gestation fetal sheep (Hawkins et al. 2000a, 2000b) or young adult sheep (Gardner et al. 2004). Apart from the obvious importance of the degree and timing of malnutrition during gestation for the development of adult hypertension and of autonomic blood pressure control, it is unclear if the different experimental nutritional interventions used have specific effects that contribute to the widely differing results between the studies. In view of the epidemiological association of poor fetal growth and adult hypertension over the normal range of birth weight (Barker, 1998), it is of particular interest to examine the development and autonomic control of fetal blood pressure (FBP) over the range of spontaneously occurring trajectories of fetal growth. In humans, twin pregnancies are a major contributor to low birth weight, but twins also follow a different trajectory of cardiovascular development independently of fetal weight (Phillips et al. 2001). Similarly, twin fetuses of pregnant ewes undernourished by 30% during the periconceptual period showed no reduced weight, but a late-gestation increase in FBP (Edwards & McMillen, 2002).

We hypothesized that variations within the normal fetal weight span are accompanied by different trajectories of development of cardiovascular autonomic control. We determined effects of spontaneously occurring weight differences in fetal sheep in relation to the baroreceptor reflex and autonomic nervous system activity at 0.76 and 0.87 gestation, i.e. before and during development of the HPA axis (Magyar et al. 1980) that is involved in cardiovascular development (Unno et al. 1999). We examined the specific influence of twin pregnancies on development of autonomic control of cardiovascular function. The baroreceptor reflex is an important contributor to short-term blood pressure control (Wood & Tong, 1999), suggesting that the baroreceptor reflex is associated with changes in short-term fetal heart rate (FHR) and variability (fHRV) measures. Thus, activity of the autonomic nervous system was estimated by fHRV analysis. Assessing fHRV to estimate the degrees of sympathetic and vagal modulation of FHR has led to valuable insights in autonomic control of cardiac function in human (Karin et al. 1993; Groome et al. 1994; Zhuravlev et al. 2002) and sheep fetuses (Kimura et al. 1996; Yu & Lumbers, 2000). Better knowledge of the effects of reduced fetal growth on development of autonomic function that is involved in cardiovascular regulation (Kimura et al. 1996) would improve our understanding of the programming of the predisposition for cardiovascular diseases in later life.

Methods

Animal husbandry

Experimental procedures were approved by the animal welfare commission of Thuringia. Time-mated German Longwool Merino × Blackheaded Mutton cross-bred ewes were bred and maintained under normal farm conditions. Ewes were kept on small weekly changing pastures from April to November, and were provided with fresh drinking water daily. Each pasture enabled the animals to seek shadow during hot weather and to go for windbreak during cold hours. Nutrition was very comprehensive and diversified and contained natural sources, for example grain, beet, fruit and cabbage. From December to March animals were kept in barns with hay and corn silage supply ad libitum and unlimited water supply.

Sixteen days before to 21 days after mating all animals were housed in a barn. Animals were individually bred after oestrus synchronization using 40 mg flurogestone acetate-based intravaginal sponges (Chronogest; Intervet). Oestrus was stimulated by intramuscular administration of 320–340 i.u. of equine chorionic gonadotrophin (Intergonan; Intervet) immediately following removal of the sponge. To detect pregnancy, maternal blood was taken at 19 days post-breeding, from the external jugular vein for determination of plasma progesterone level using a direct chemiluminescent competitive immuno-assay based on an acridinium ester-labelled mouse monoclonal antiprogesterone antibody (ACS: Centaur Progesterone Assay; Chiron Diagnostics) (Muller et al. 2003).

Surgical procedure

The pregnant sheep were brought into the animal facilities at least five days before surgery and kept in rooms with controlled light–dark cycles (12 h light and 12 h dark: lights off at 6.00 p.m. and lights on at 6.00 a.m.), as well as air conditioning (adjusted to 18°C and 50% relative air humidity). Hay or hay cubes and water were provided ad libitum. After food withdrawal for 24 h, body weight of the ewes was determined and surgery was performed in 18 fetuses at 109 ± 1 (mean ±s.e.m.) days of gestation (dGA, term 150 dGA, 0.73 gestation), and 17 fetuses at 125 ± 3 dGA (0.83 gestation). Ewes were pretreated with 1 g of ketamine (Ketamin 10, Atarost, Germany) i.m. and intubated under 4% halothane (Fluothane, Zeneca, Germany) using a face mask. Anaesthesia was maintained with 1.0–1.5% halothane in 100% oxygen. Ewes were instrumented with catheters inserted into the common carotid artery for blood sampling and into the external jugular vein for postoperative administration of drugs. Laparotomy and hysterotomy were performed. In twin pregnancies, the usually only slightly smaller fetus was identified by comparing the vertex-to-nose length and the interorbital span of the fetal heads through the unopened uterus. This fetus was removed to provide the same intrauterine conditions for all fetuses during instrumentation. The removed fetus was killed immediately, with an overdose of intracardiac sodium pentobarbital (Narcoren, Merial, Germany) while still under general anaesthesia with halothane. Polyvinyl catheters (Rüschelit, Rüsch, Germany) were inserted into the left fetal common carotid artery for blood pressure acquisition and blood sampling for blood gas measurements, and into the fetal left external jugular vein for sodium nitroprusside (SNP) and phenylephrine (PE) administration. Another catheter was placed in the amniotic cavity for amniotic pressure (AMN) recordings to correct FBP for hydrostatic pressure. Wire electrodes (LIFYY, Metrofunk Kabel-Union, Berlin, Germany) were implanted into the left suprascapular muscles, muscles of the right shoulder and in the cartilage of the sternum to record the electrocardiogram (ECG), and into the uterus to monitor myometrial activity. Following closure of the uterus, cables and catheters were exteriorized and the maternal abdomen closed.

After surgery, ewes were returned to a metabolic cage and allowed at least 5 days of postoperative recovery. Ewes and fetuses received 0.5 g ampicillin (Ratiopharm, Germany) i.v. and into the amniotic sac, respectively, twice a day during the first three postoperative days. Metamizol (Arthripur, Atarost, Germany, dose range of 30–50 mg kg−1) was administered i.v. daily as analgesic for at least three days. All catheters were maintained patent via a continuous pump infusion of heparin at 15 i.u. ml−1 in 0.9% NaCl solution delivered at a rate of 0.5 ml h−1.

Experimental protocol

Fetal and maternal arterial blood samples were taken daily at 8.00 a.m. for measurement of blood gases and pH (ABL 600, Radiometer, Copenhagen, Denmark). Measurements were corrected to 39°C. FBP and AMN were measured using calibrated pressure transducers (Braun, Melsungen, Germany); 48 h before the start of the experiments continuous baseline recordings of FBP, AMN, ECG and myometrial activity were initiated. Sensitivity (linear FHR/FBP slopes) and range (FHR/FBP limits) of the baroreceptor reflex were explored during pharmacologically induced acute changes of FBP at 0.76 and 0.87 gestation. FBP was decreased or increased by stepwise infusion of SNP (Sigma-Aldrich, Deisenhofen, Germany, 10 μg ml−1, infused at a rate of 0.2–0.8 ml min−1 increased in steps of 0.1 ml min−1) or PE (Sigma-Aldrich, 25 μg ml−1, infused at a rate of 0.2–0.5 ml min−1 increased in steps of 0.1 ml min−1) to the fetal jugular vein. Each new infusion rate was maintained constant over 120 s in the 0.87 gestation fetuses, and 60 s in the 0.76 gestation fetuses, to avoid potential volume effects on FBP in the younger fetuses. Nevertheless, FBP reached a plateau at each infusion rate at both gestational ages. A break of at least half an hour was permitted between the SNP and PE experiments to allow normalization of FBP.

One day after the experiment, ewes and their fetuses were killed by maternal i.v. injection of a sodium pentobarbital overdose (Narcoren, Merial, Germany). Fetal body weights as well as fetal brain, heart and liver weights were determined, and relative organ weights as well as the brain/liver ratio were calculated.

Data recording and analysis

FBP, AMN, ECG and uterine electromyogram (EMG) were amplified (Gould, Valley View, OH, USA), digitized using a 16-channel A/D board (DT 2801F, Data Translation, Marlborough, MA, USA, sampling rate of ECG 1024 Hz, sampling rate of FBP, AMN and EMG 64 Hz), and continuously stored on hard discs of a PC throughout the experimental period. Mean FBP was estimated from the raw blood pressure signal by a sliding rectangular window averaging the blood pressure over 1 s periods (Watisa, Institute of Medical Statistics, Computer Sciences and Documentation, Friedrich Schiller University, Jena, Germany). This procedure corresponds to calculating mean arterial blood pressure by determining the area under the pressure wave form averaged over the cardiac cycle. Age-related FBP and FHR were calculated from recordings obtained between midnight and 6.00 a.m. when the animals were undisturbed by animal care and experiments. Periods during uterine contractures detected by uterine electromyography that occur approximately every 20 min at the gestational ages examined were not analysed. FBP was corrected for AMN. For FHR and fHRV analysis, beat-to-beat intervals were sequentially detected and triggered from R peaks in the ECG with a precision of ±0.49 ms.

In order to compare autonomic and cardiovascular development in relation to fetal weight, all fetuses were allocated to a group with lower or higher body weight at each gestational age, i.e. four groups were established using the so-called K-means cluster analysis with K= 4 (SPSS 11.5, SPSS Inc., Chicago, IL, USA).

FHRV was analysed over 300 s before testing the baroreceptor reflex. All animals showed stable records and were in an undisturbed condition during this time period (no food intake, no irregular external noises or any further stressors). Artifacts were visually controlled for and removed manually (Watisa). Two animals at 0.76 gestation had to be excluded due to numerous ECG artifacts. The resulting instantaneous R–R intervals sequence was linearly interpolated at the equidistant sample rate of 1000 Hz and resampled at 20 Hz for calculation of fHRV measures. Stationarity of fHRV is a prerequisite for fHRV analysis and proven by showing that R–R intervals are normally distributed, i.e. statistical properties of the signal are the same in every part of the signal. The condition of stationarity was tested and confirmed in all R–R intervals, by means of a Kolmogorov–Smirnov test for normality of the mean FHR and SDNN distributions with Lilliefors correction for significance (SPSS 11.5) (Mansier et al. 1996). We determined the following measures of fHRV using Matlab 6.1, R13 (The MathWorks, Natick, MA, USA) according to the recommendations of the Task Force (1996): SDNN, HRVtri and LF band spectral power (0.04–0.2 Hz) were calculated as fHRV time and frequency domain measures describing long-term variability that is influenced by both sympathetic and vagal activity (Groome et al. 1994). RMSSD and HF band spectral power (0.2–2 Hz) were calculated as fHRV time and frequency domain measures describing the short-term variability that is influenced by vagal activity (Groome et al. 1994). The LF/HF ratio was calculated as an indicator of sympathovagal balance (Troeger et al. 2003). To ensure that no aliasing occurred during the linear interpolation and re-sampling described above, spectral analysis of instantaneous R–R intervals sequences sampled at 1000 Hz was performed. The part of spectral power at frequencies higher than 2 Hz was less than 1% of total power.

Baselines of FBP and FHR prior to the start of SNP or PE infusion were determined as an average of FBP and FHR over 300 s. Baroreceptor reflex response curves were evaluated by summarizing FBP changes in 2.0 mmHg intervals, and assigning the respective FHR values. Linear slopes between baselines and the upper FHR/lower FBP limit during SNP infusion or the lower FHR/upper FBP limit during PE infusion was calculated. The limit of baroreceptor reflex regulation was considered as the point at which FHR remained stable, i.e. did not change more than 5 min−1 per 2.0 mmHg FBP change. The FBP range of baroreceptor reflex regulation was determined as the range over which the FBP is controlled during the baroreceptor reflex challenge, i.e. as the difference between the baseline and the maximal or minimal FBP value achieved.

Statistical analysis

Correlations of fetal body weights to baseline FBP and fHRV-derived measures were calculated using the Spearman-Rho correlation coefficient for non-parametric values (SPSS 11.5). Differences in fetal weights and numbers, maternal weights, breeding season, blood gases, FBP, FHR, between fHRV-derived measures slopes and limits of baroreceptor reflex between fetal weight groups and gestational ages were tested for significance by Mann–Whitney U tests. P values were adjusted using the Bonferroni–Holm method. Significance was assumed at a P value < 0.05. All values are given as mean ±s.e.m. except in Fig. 1.

Figure 1.

Relationship of fetal body weight and FBP
A is without and B with display of fetal weight groups. Each point represents one fetus at 0.76 gestation (squares) and 0.87 gestation (circles). Open symbols reveal low-weight and filled symbols normal-weight fetuses identified by cluster analysis. Grey symbols are the means of each weight group ±s.d.*P < 0.05 for fetuses at 0.76 gestation. Please note that correlation was calculated over all fetuses at the respective age, i.e. before establishing the weight groups.

Results

Fetal body weight

Fetal body weights correlated to FBP at 0.76 but not at 0.87 gestation (P < 0.05, R=−0.59, Fig. 1A). Grouping of fetal weights by cluster analysis to quantify the effect of fetal weight on cardiovascular regulation resulted in two groups of fetuses at each gestational age, with weights of 1505 ± 31 g (n= 7) and 2076 ± 64 g (n= 11, P < 0.01) at 0.76 gestation and 3244 ± 92 g (n= 11) and 4507 ± 211 g (n= 6, P < 0.01) at 0.87 gestation (Table 1, Fig. 1B). All fetuses were within the normal weight range defined by the 5th and 95th percentile calculated from 38 fetuses of our breeding at 0.76 (1192 and 2303 g, respectively) and of 43 fetuses at 0.87 gestation (1750 g and 5008 g, respectively, own unpublished results). Consequently, low-weight fetuses (LW) did not meet the definition of intrauterine growth retardation. We termed the fetuses with the higher weights ‘normal weight’ (NW), as under natural conditions sheep fetuses do not gain overweight but they may have lower weights for several reasons discussed below.

Table 1.  Fetal body and organ weights at 0.76 and 0.87 gestation in relation to fetal weight
 0.76 gestation0.87 gestation
LW (n= 7)NW (n= 11)LW (n= 11)NW (n= 6)
  1. Organ weights are also given relative to body weight. LW, low-weight fetuses; NW, normal-weight fetuses. Values are means ±s.e.m. *P < 0.05; **P < 0.01 in comparison to NW at the same gestational age; †P < 0.05; ‡P < 0.01 in comparison to the corresponding weight at 0.76 gestation.

Gestational age (days)114 ± 0  115 ± 1  128 ± 1‡ 132 ± 1*‡
Number of offspring2.0 ± 0.31.7 ± 0.21.5 ± 0.2 1.0 ± 0.0†
Fetal body weight (g)1495 ± 33** 2076 ± 64   3227 ± 100**‡4507 ± 211‡
Fetal heart weight (g)14 ± 1 16 ± 1 25 ± 1† 30 ± 3‡
Relative heart weight (%)  1.0 ± 0.0** 0.7 ± 0  0.8 ± 0.0†0.7 ± 0.1
Fetal brain weight (g)33 ± 1 34 ± 2 46 ± 2‡51 ± 2‡
Relative brain weight (%)  2.2 ± 0.0** 1.6 ± 1  1.4 ± 0.1*‡ 1.2 ± 0.1‡
Fetal liver weight (g)86 ± 8 96 ± 5 126 ± 12†168 ± 18‡
Relative liver weight (%)5.9 ± 0.44.7 ± 0.3 4.0 ± 0.2‡3.8 ± 0.4
Brain/liver ratio0.37 ± 0.040.38 ± 0.050.37 ± 0.02 0.32 ± 0.03

Fetal values of arterial pH, Pmath formula and Pmath formula of the four groups were within the physiological range during postsurgical recovery and on the day of experiment (Table 3). Therefore, the small differences between the groups did not have pathophysiological relevance.

Table 3.  Fetal arterial pH and blood gases at the day of experiment
 0.76 gestation0.87 gestation
LW (n= 7)NW (n= 11)LW (n= 11)NW (n= 6)
  1. LW, low-weight fetuses; NW, normal-weight fetuses. Values are means ±s.e.m.*P < 0.05 in comparison to the corresponding weight at 0.76 gestation.

pH value 7.31 ± 0.02 7.34 ± 0.01 7.37 ± 0.01* 7.34 ± 0.02
P math formula (mmHg)42.6 ± 3.043.0 ± 1.345.1 ± 1.339.7 ± 2.2
P math formula (mmHg)21.8 ± 1.727.7 ± 1.226.4 ± 1.722.3 ± 1.5*

The body weights of the ewes of NW and LW fetuses were similar (66.0 ± 6.1 versus 71.8 ± 4.9 kg). Absolute fetal heart, brain and liver weights did not differ between LW and NW fetuses at both gestational ages. Relative heart and brain weights were higher in LW than in NW fetuses at 0.76 gestation (P < 0.01, Table 1) and relative brain weight was higher at 0.87 gestation (P < 0.05, Table 1). The absence of differences in the brain/liver ratio supports the notion that the LW fetuses did not meet the classic definition of intrauterine growth restriction. Absolute heart, brain and liver weights increased in LW and NW fetuses from 0.76 to 0.87 gestation (at least P < 0.05, Table 1). Relative brain weight decreased in NW fetuses (P < 0.01, Table 1) and relative heart, brain, and liver weights decreased in LW fetuses (at least P < 0.05, Table 1).

Effects of twin pregnancies The percentage of twins tended to be higher in LW (5 out of 7 or 71.4%) than in NW (5 out of 11 or 45.5%) fetuses at 0.76 gestation and was higher at 0.87 gestation (P < 0.05, LW: 5 out of 11 or 45.5%; NW: 0 out of 6). No differences in fetal body weights and in absolute or relative organ weights were found at 0.76 gestation. At 0.87 gestation, body, brain (P < 0.05) and, in tendency, liver weights (P < 0.07) of the singletons were higher than in twins (Table 2). Absolute body, brain, heart, and liver weights increased in both singletons and twins from 0.76 to 0.87 gestation (P < 0.01, Table 2). Relative brain (P= 0.07) and liver (P < 0.05) weight decreased in singletons but not in twins (Table 2).

Table 2.  Fetal body and organ weights at 0.76 and 0.87 gestation in relation to number of fetuses
 0.76 gestation0.87 gestation
Singletons (n= 7)Twins (n= 11)Singletons (n= 12)Twins (n= 5)
  1. Organ weights are also given relative to body weight. Values are means ±s.e.m.§P= 0.07; *P < 0.05 in comparison to singletons at the same gestational age; +P= 0.07; †P < 0.05; ‡P < 0.01 in comparison to the corresponding number of fetuses at 0.76 gestation.

Gestational age (days)115 ± 1 114 ± 0 130 ± 1‡ 128 ± 1‡
Fetal body weight (g)1875 ± 1161822 ± 1073930 ± 214‡ 3074 ± 94*‡
Fetal heart weight (g)15 ± 215 ± 130 ± 2‡ 24 ± 0†
Relative heart weight (%)1.0 ± 0 1.0 ± 0 1.0 ± 0 1.0 ± 0 
Fetal brain weight (g)32 ± 134 ± 251 ± 1‡  42 ± 0*† 
Relative brain weight (%)2.0 ± 0 2.0 ± 0 1.0 ± 0+1.0 ± 0 
Fetal liver weight (g)98 ± 289 ± 6160 ± 11‡116 ± 9§
Relative liver weight (%)5.0 ± 0 5.0 ± 0 4.0 ± 0† 4.0 ± 0 
Brain/liver ratio 0.34 ± 0.02 0.40 ± 0.040.33 ± 0.02 0.37 ± 0.02

Fetal body weight and cardiovascular development

FBP was higher in LW than in NW fetuses at 0.76 gestation (P < 0.01, Table 4). At 0.87 gestation, FBP was not different in LW and NW fetuses (Table 4). FBP increased in NW but not in LW fetuses between 0.76 and 0.87 gestation (P < 0.01, Table 4). No differences in FHR were found between LW and NW fetuses at 0.76 and 0.87 gestation (Table 6). FHR decreased in LW but not in NW fetuses between 0.76 and 0.87 gestation (P < 0.05, Table 6).

Table 4.  FBP and FHR changes during baroreceptor reflex challenges in relation to fetal weight
 0.76 gestation0.87 gestation
LW (n= 7)NW (n= 11)LW (n= 11)NW (n= 6)
  1. LW, low-weight fetuses; NW, normal-weight fetuses. Values are means ±s.e.m.For baseline FHR see Table 6. *P < 0.05; **P < 0.01 in comparison to NW at the same gestational age; §P= 0.06; †P < 0.05; ‡P < 0.01 in comparison to the corresponding weight at 0.76 gestation.

FBP (mmHg)  49 ± 3** 39 ± 148 ± 1 51 ± 1‡
SNP response
 Slope−10 ± 6 −6 ± 2−4 ± 2−3 ± 1
 FBP range of baroreceptor reflex regulation (mmHg)−6 ± 2−6 ± 2−11 ± 2 −14 ± 2§
 Maximal FHR change (min−1)  23 ± 12  31 ± 1130 ± 8 31 ± 6
PE response
 Slope−10 ± 3*−3 ± 1 −2 ± 0‡−2 ± 0
 FBP range of baroreceptor reflex regulation (mmHg)  8 ± 3 15 ± 1 22 ± 2‡ 25 ± 3†
 Maximal FHR change (min−1)−44 ± 7 −44 ± 6 −47 ± 3 −58 ± 6
Table 6.  fHRV measures in relation to fetal weight
 0.76 gestation0.87 gestation
LW (n= 7)NW (n= 9)LW (n= 11)NW (n= 6)
  1. LW, low-weight fetuses; NW, normal-weight fetuses. Values are means ±s.e.m.*P < 0.05; **P < 0.01 in comparison to NW at the same gestational age; †P < 0.05; ‡P < 0.01 in comparison to the corresponding weight at 0.76 gestation.

FHR (min−1)190 ± 7  180 ± 6  173 ± 3† 181 ± 5  
fHRV time domain
 SDNN (ms)9.8 ± 1.67.6 ± 1.09.8 ± 0.813.7 ± 2.0 
 RMSSD (ms)7.0 ± 1.54.3 ± 0.7 5.8 ± 1.0* 9.8 ± 0.6‡
 HRVtri2.9 ± 0.42.3 ± 0.33.0 ± 0.34.4 ± 0.5
fHRV frequency domain
 LF (min−2)12.8 ± 6.4 9.6 ± 3.09.0 ± 1.421.1 ± 7.5 
 HF (min−2)3.0 ± 0.82.0 ± 0.7  2.2 ± 0.5** 6.8 ± 1.2‡
 LF/HF5.1 ± 1.58.8 ± 3.07.8 ± 3.12.9 ± 1.0

The higher baseline FBP in LW fetuses at 0.76 gestation was accompanied by a shift of the baroreceptor reflex response curve to the right (Fig. 2). In addition, we found a higher slope and, thus, baroreceptor reflex sensitivity in LW than in NW fetuses during the hypertensive challenge (P < 0.05) and, in tendency, during the hypotensive challenge (Table 4, Fig. 2). The higher baroreceptor reflex sensitivity disappeared at 0.87 gestation (P < 0.01, Table 4, Fig. 2). The baroreceptor reflex sensitivity did not show developmental changes in NW fetuses. The FBP range of baroreceptor reflex regulation and FHR response to the FBP increase and decrease did not differ between the weight groups at both gestational ages. The FBP range of baroreceptor reflex regulation increased during the hypertensive challenge in LW and NW fetuses with gestational age (P < 0.05), and in tendency during the hypotensive challenge (Table 4). The FHR response to the FBP increase or decrease did not show developmental changes in NW and LW fetuses.

Figure 2.

Baroreceptor reflex responses in relation to fetal body weight and gestational age
Large symbols represent baselines. Thin lines represent low-weight and bold lines normal-weight fetuses. Values are means ±s.e.m.

Effects of twin pregnancies FBP and FHR did not differ between singletons and twins at either gestational age. FBP increased in singletons but not in twins between 0.76 and 0.87 gestation (P < 0.01, Table 5). FHR did not change between the gestational ages. Fetal body weights of twins correlated negatively to FBP at 0.76 but not 0.87 gestation, and singletons showed a strong tendency of correlation (P= 0.05, R=− 0.60, Fig. 3). At 0.76 gestation, the baroreceptor reflex sensitivity during the hypotensive challenge was lower and the FHR increase in response to the FBP decrease less pronounced in twins than in singletons (P < 0.05, Table 5). At 0.87 gestation, no differences in baroreceptor reflex regulation were found between singletons and twins, as between NW and LW fetuses. Contrary to the NW fetuses, the baroreceptor reflex sensitivity in response to the hypotensive challenge decreased in singletons but not in twins between 0.76 and 0.87 gestation (P < 0.05, Table 5). Similar to the NW and LW fetuses, the range of baroreceptor reflex regulation during the hypertensive and hypotensive challenges was, at least in tendency, higher in singletons and twins at 0.87 versus 0.76 gestation (P < 0.05), while the FHR responses to the FBP increases or decreases did not show any developmental changes (Table 5).

Table 5.  FBP and FHR changes during baroreceptor reflex challenges in relation to number of fetuses
 0.76 gestation0.87 gestation
Singletons (n= 7)Twins (n= 11)Singletons (n= 12)Twins (n= 5)
  1. Values are means ±s.e.m. For baseline FHR see Table 7. *P < 0.05 in comparison to singletons at the same gestational age; †P < 0.0; ‡P < 0.01 in comparison to the corresponding number of fetuses at 0.76 gestation.

FBP (mmHg)41 ± 245 ± 2 50 ± 1‡48 ± 2
SNP response
 Slope−12 ± 4 −2 ± 1*−3 ± 1†−5 ± 2
 FBP range of baroreceptor reflex regulation (mmHg)−5 ± 1−7 ± 1 −13 ± 2† −9 ± 3
 Maximal FHR change (min−1)40 ± 8  7 ± 4*31 ± 5 29 ± 11
PE response
 Slope−5 ± 2−6 ± 2−2 ± 0 −2 ± 0
 FBP range of baroreceptor reflex regulation (mmHg)13 ± 212 ± 223 ± 222 ± 2
 Maximal FHR change (min−1)−48 ± 8 −41 ± 5 −53 ± 3  −46 ± 6 
Figure 3.

Relation of fetal body weight and FBP in the subgroups of singletons and twins
Each point represents one fetus. Regression lines are fitted for P < 0.05.

Fetal body weight and fHRV

At 0.76 gestation, neither significant correlations between fetal body weights and fHRV measures over all fetuses nor differences of fHRV measures between LW and NW fetuses were found (Table 6, Fig. 4). At 0.87 gestation, fetal body weights of all fetuses were directly correlated to the fHRV measures RMSSD and HF band spectral power that reflect vagal modulations of FHR (P < 0.01, both R= 0.73, Fig. 4). Consequently, these measures were lower in LW than in NW fetuses at this age (at least P < 0.05), and were increased in NW fetuses at 0.87 compared to 0.76 gestation (P < 0.01, Table 6).

Figure 4.

Correlation of fHRV measures RMSSD and HF band spectral power to fetal weights
0.76 gestation, • 0.87 gestation. *P < 0.05. Please note that correlation was calculated over all fetuses at the respective age, i.e. before establishing the weight groups. Regression lines are fitted for P < 0.05.

Effects of twin pregnancies There were no differences of fHRV measures between singletons and twins at 0.76 or 0.87 gestation, or between 0.76 and 0.87 gestation except an increased HRVtri in singletons at 0.87 versus twins and versus singletons at 0.76 gestation (at least P < 0.05, Table 7). Similar to the group of all fetuses at 0.76 gestation, the subgroups of singletons and twins did not show significant correlations between fetal body weights and fHRV measures (Fig. 5). At 0.87 gestation, the subgroup of singletons but not of twins showed a direct correlation between fetal body weights and the fHRV measures RMSSD and HF band spectral power, similar to the group of all fetuses (at least P < 0.05, R= 0.51 and R= 0.46, Fig. 5). In addition, fetal weights of singletons showed a direct correlation to HRVtri, a measure of both sympathetic and vagal modulation of FHR (P < 0.05, R= 0.40, figure not shown).

Table 7.  fHRV measures in relation to number of fetuses
 0.76 gestation0.87 gestation
Singletons (n= 6)Twins (n= 10)Singletons (n= 12)Twins (n= 5)
  1. Values are means ±s.e.m.*P < 0.05 in comparison to singletons at the same gestational age; ‡P < 0.01 in comparison to the corresponding number of fetuses at 0.76 gestation.

FHR (min−1)183 ± 4  185 ± 7  178 ± 3  170 ± 4  
fHRV time domain
 SDNN (ms)8.4 ± 1.38.7 ± 1.412.2 ± 1.2 8.7 ± 1.1
 RMSSD (ms)6.4 ± 1.85.0 ± 0.97.9 ± 1.05.5 ± 0.9
 HRVtri2.4 ± 0.32.6 ± 0.4 3.9 ± 0.4‡ 2.6 ± 0.3*
fHRV frequency domain
 LF (min−2)8.6 ± 3.312.4 ± 5.1 15.5 ± 4.0  7.9 ± 2.5 
 HF (min−2)3.2 ± 1.22.0 ± 0.64.4 ± 1.02.5 ± 0.7
 LF/HF4.3 ± 1.18.9 ± 2.87.0 ± 2.93.9 ± 1.1
Figure 5.

Correlation of fHRV measures RMSSD and HF band spectral power to fetal weights in the subgroups of singletons and twins
Each point represents one fetus. Regression lines are fitted for P < 0.05.

Discussion

The present study shows that differences in the development of the cardiovascular system and autonomic control of cardiovascular and cardiac function are associated with fetal weight within the normal weight span during the last third of gestation in sheep. None of the fetuses was below the 5th weight percentile or met the definition of intrauterine growth restriction (Wollmann, 1998). At 0.76 gestation, FBP was negatively correlated to fetal weight. FBP values in LW fetuses were already at the level that NW fetuses reached at 0.87 gestation. The increased FBP in conjunction with a trend to a higher FHR suggests sympathetic activation in these fetuses. In parallel, the baroreceptor reflex sensitivity was increased in LW versus NW fetuses at 0.76 gestation as a sign of immaturity of circulatory control (see below). As one of the most striking results, RMSSD and HF band spectral power, reflecting vagal modulation of FHR, were linearly and directly correlated to the fetal weight at 0.87 gestation, i.e. after onset of maturation of vagal FHR control, suggesting delayed maturation of vagal activity in LW fetuses. The majority of LW fetuses were twins but weight differences between singletons and twins were absent at 0.76 and minor at 0.87 gestation. In agreement with this, and in contrast to LW fetuses, twin fetuses showed no signs of sympathetic activation regarding blood pressure control at 0.76 and 0.87 gestation as well as no definite signs of an altered maturation of autonomic FHR control.

Organ growth

The differences in relative organ weights between LW and NW fetuses at both gestational ages suggest different trajectories of fetal growth. Variations of the trajectories of fetal growth are considered aetiologically important for development of cardiovascular diseases in adult life (Barker, 1998; McMillen & Robinson, 2005). Primary growth of the carcass during the last third of gestation is a normal feature of development reflected in the exponential growth curve found in all species. The lower relative heart, brain and liver weights at 0.87 compared to 0.76 gestation in LW fetuses reflect a relatively higher carcass mass at 0.87 gestation, and suggest primary carcass growth. In contrast, only relative brain weight decreased in NW fetuses between 0.76 and 0.87 gestation. Comparing LW and NW fetuses, the higher relative heart and brain weights in LW than in NW fetuses at 0.76 gestation and the higher relative brain weight in LW than in NW fetuses at 0.87 gestation reflect a certain degree of growth dysproportionality. Asymmetry in growth is thought to reflect a deficiency in nutrient supply, particularly during the last third of gestation (Wollmann, 1998). In the present study, deficiency in fetal nutrient supply must have been moderate as we found only slight dysproportionality of growth in LW fetuses and no typical signs of asymmetric growth restriction such as an increased brain/liver-ratio.

Low fetal nutrient supply is a fetal stressor independent of the cause of malnutrition (Barker, 1998; McMillen & Robinson, 2005). Major contributors to low fetal weight are the balance of nutrients, maternal size, number of fetuses and placental nutrient exchange (McMillen & Robinson, 2005). While we did not standardize nutrients, all sheep had access to the same nutrients and did not show seasonal variations in fetal weight. Moreover, variation in food supply produces a relatively minor effect on birth weight compared to the constraints of maternal size and placental function (Harding, 2001). Maternal weight did not differ between the groups in our study. The higher proportion of twin pregnancies in the LW groups is in good agreement with the human situation (Morley et al. 2003) and experimental studies in sheep (for review see Luther et al. 2005). Reduced fetal weights, however, cannot be explained by twin pregnancies alone, as we did not find fetal weight differences in singletons and twins at 0.76 gestation. Placental function is crucial for fetal nutritional status, and may not only explain the occurrence of LW fetuses, but also differences in growth in singleton and twin pregnancies (Luther et al. 2005; McMillen & Robinson, 2005). We cannot determine placental dysfunction as a contributor to low fetal weight, since we did not measure placental weight and umbilical flow.

Effects of twin pregnancies While twin pregnancies contributed above average to the LW groups, weight differences between singletons and twins were absent at 0.76 gestation and modest with lower body, brain and liver weights at 0.87 gestation, demonstrating that twin pregnancy belongs to normality in sheep and is not a major malnutritive condition per se. Nutrient supply in twin pregnancies seems to only become limited in the period of exponential growth towards the end of gestation. It has been shown in sheep that if undernutrition is prolonged during late pregnancy, fetal growth is particularly compromised in twin pregnancies (Luther et al. 2005).

Cardiovascular development

FBP increase, FHR decrease and decrease of baroreceptor reflex sensitivity over the last third of gestation are distinct features of cardiovascular development in sheep (Blanco et al. 1988; Crowe et al. 1995; Unno et al. 1999; Shinozuka et al. 2000). The elevated FBP in LW fetuses at 0.76 gestation and the absent FBP increase between 0.76 and 0.87 gestation in the LW fetuses suggest that LW fetuses already reached FBP values at 0.76 gestation similar to those reached in NW fetuses at 0.87 gestation. Earlier studies also failed to show a developmental increase of FBP during late gestation in growth-restricted fetuses between 0.80 and 0.93 gestation (Louey et al. 2000), or in fetuses with moderate weight reduction between 0.77 and 0.86 gestation (Crowe et al. 1995). Increased FBP in LW fetuses at 0.76 gestation reflects early maturation of sympathetic activity (Segar et al. 1994), altered vascular reactivity (Ozaki et al. 2000) or increased cardiac contractility (McMillen & Robinson, 2005). Maternal nutrient restriction by 50% for 12 days prior to and 70 days after conception resulted in a blunted endothelium-dependent and non-endothelium-dependent vasodilatation at 130 days' gestation (Ozaki et al. 2000). Increased sympathetic activity and increased levels of circulating catecholamines are typical features of intrauterine growth restriction, both of which have been causally linked to programming of cardiovascular diseases in later life (Lumbers et al. 2001). Developmental increase of FBP also parallels maturation of the HPA axis (Blanco et al. 1988; Crowe et al. 1995; Unno et al. 1999; Shinozuka et al. 2000; Hawkins et al. 2000a). The physiological development of the cardiovascular system is dependent on fetal adrenal function (Unno et al. 1999). Glucocorticoids induce myocardial growth (Rudolph et al. 1999), enhance vascular sensitivity to endothelin, and decrease nitric oxide synthase activity (Tangalakis et al. 1992), and, hence, may contribute to the developmental increase of peripheral resistance and cardiac output. Insufficient fetal nutrient supply at different stages of pregnancy alters responsiveness of the fetal HPA axis in late-gestation sheep (Hawkins et al. 2000a; Edwards et al. 2001; Edwards & McMillen, 2002). Increased activity of the HPA axis was associated with an increase of FBP and vascular sensitivity to angiotensin II (Edwards & McMillen, 2001). Alternatively, there may be an enhanced central action of angiotensin II, resulting in enhanced stimulation of sympathetic tone and concomitant inhibition of vagal tone as found in the present study. The specific impact of early undernutrition on the programming of the HPA axis appears to be dependent on the degree and duration of undernutrition. The periconceptual and/or the preimplantational periods in particular are critical windows during which maternal undernutrition may act to alter the set-point of the function of the HPA axis (McMillen & Robinson, 2005).

We have found a developmental FHR decrease in LW but not in NW fetuses. This decrease does not seem to be vagally mediated, since vagal modulation of FHR was less pronounced in LW than in NW fetuses at 0.87 gestation (see discussion of fHRV results below). It is more likely to be due to an increased sympathetic activity in LW fetuses at 0.76 gestation. The absent developmental FHR decrease in NW is in agreement with several other studies that failed to show the developmental FHR decrease (Walker et al. 1978; Crowe et al. 1995), probably because ontogenetic changes of baseline FHR are very slight.

The baroreceptor reflex is an important contributor to short-term blood pressure control (Wood & Tong, 1999). In fetal sheep, this reflex is functional relatively early. Spontaneous carotid and aortic baroreceptor discharges can be recorded from 0.57 gestation onwards (Blanco et al. 1988), and the efferent limb of the baroreceptor reflex is functional from 0.40 gestation (Born et al. 1956). Baroreceptor reflex sensitivity decreases during development (Walker et al. 1983; Blanco et al. 1988; Segar, 1997), and this decrease is mediated by maturation of both the afferent (vagal) (Blanco et al. 1988) and efferent (sympathetic) limb of the baroreceptor reflex (Segar, 1997). Our results on fHRV suggesting vagal maturation between 0.76 and 0.87 gestation (see discussion of fHRV results below) are in good agreement with the notion that the afferent part of the baroreceptor reflex contributes to the developmental decrease of baroreceptor reflex sensitivity (Blanco et al. 1988). Interestingly, we did not find a developmental decrease of baroreceptor reflex sensitivity in NW fetuses, in spite of the maturation of vagal modulation of FHR, suggesting that maturation of autonomic baroreceptor reflex and cardiac control are not strictly temporally concomitant, at least as far as the time span of 0.76–0.87 gestation is concerned. Our results in this study and the previous studies cited above suggest an earlier maturation of the baroreceptor reflex than of vagal cardiac control. The period of development examined may also be too short to reveal a partial or even complete maturation of the baroreceptor reflex, given the ongoing development of baroreceptor reflex sensitivity into the postnatal period (Blanco et al. 1988; Segar, 1997). The higher baroreceptor reflex sensitivity in LW versus NW fetuses at 0.76 gestation, combined with ‘catch-up’ maturation of the baroreceptor reflex until 0.87 gestation, may reflect immaturity of the baroreceptor reflex and higher sympathetic tone of vascular control in LW fetuses at 0.76 gestation. ‘Catch-up’ maturation of the baroreceptor reflex between 0.76 and 0.87 gestation, and delayed maturation of vagal modulation of FHR (see discussion of fHRV results below) in LW fetuses suggest a different trajectory or a general delay of maturation of the autonomic nervous system, maintining the order of maturation of autonomic baroreceptor reflex control followed by maturation of autonomic cardiac control that was found in NW fetuses. The shift of the baroreceptor reflex response curve to the right at 0.76 gestation reflects resetting of the baroreceptors due to the higher FBP in the LW fetuses.

The FBP range of baroreceptor reflex regulation increased in both LW and NW fetuses between 0.76 and 0.87 gestation, mainly due to an increase of the upper limit of baroreceptor reflex regulation. This suggests a more effective regulation of FBP changes with increasing gestational age independently of fetal weight. The similar FBP range of baroreceptor reflex regulation in both LW and NW fetuses at the respective gestational ages implies that the differences in baroreceptor reflex sensitivity between LW and NW fetuses at 0.76 gestation do not severely affect blood pressure regulation. In contrast to the maturation of the FBP range of baroreceptor reflex regulation, FHR changes in response to baroreceptor activation were independent of gestational age and fetal weight.

Effects of twin pregnancies Although twin fetuses contributed above average to the LW groups, twins did not show an increased FBP compared to singletons at 0.76 gestation as did the LW versus NW fetuses. However, similar to LW fetuses, they failed to show a developmental FBP increase, suggesting that the mechanisms that lead to an early FBP increase affect twins to a lesser extent than LW fetuses. Maturation of FHR followed a similar trajectory of development in singleton and twin fetuses, resembling that in NW fetuses.

Similar to the NW and LW fetuses, the FBP range of baroreceptor reflex regulation did not differ between singletons and twins at the respective gestational ages, and increased in both singletons and twins between 0.76 and 0.87 gestation, mainly due to an increase of the upper limit of baroreceptor reflex regulation. The similar FBP range of baroreceptor reflex regulation in singletons and twins at the respective gestational ages implies that the slight differences in baroreceptor reflex sensitivity and FHR response between singletons and twins at 0.76 gestation do not profoundly affect the ability to cope with FBP changes. The slightly different trajectories of cardiovascular development in singleton and twin fetuses are in part due to the heterogeneous distribution of body weight in these groups compared to the groups of LW and NW fetuses. This may explain why both twins and singletons show a correlation between body weight and FBP at 0.76 gestation.

fHRV

Fetal body weight and fHRV measures, RMSSD and HF band spectral power, reflecting vagal modulation of FHR, were linearly and directly correlated with fetal weight at 0.87 gestation, suggesting a strong association of fetal body weight and vagal activity at this age. This correlation was absent in NW and LW fetuses as well as in singletons and twins at 0.76 gestation. The reason for this is the high variability of RMSSD and HF band spectral power values in both NW and LW fetuses at 0.76 gestation. The variability of RMSSD and HF band spectral power values apparently reflects immature vagal control of FHR. This finding suggests that development of vagal modulation of FHR is reflected rather in regularization of vagal fHRV control than in a mere increase of vagal activity. In agreement with this, Karin et al. (1993) showed a developmental decrease of fHRV power at frequencies from 0.2 to 1.0 Hz in human fetuses between 0.6 gestation and term, which was interpreted as evolution of a stable and mature activity of the autonomic nervous system. Yum & Kim, (2003) observed instabilities of FHR in human fetuses at 0.69 gestation that gradually diminished with development up to 0.93 gestation, reflecting unstable rather than absent autonomic control of FHR earlier in gestation. This phenomenon might be particularly reflected in fHRV measures that quantify short-term fluctuations of FHR in a time domain, such as the RMSSD. Thus, ‘regularization’ instead of a pure increase of activity seems to be a typical feature of maturation of the autonomic cardiac control.

RMSSD and HF band spectral power increased in NW but not LW fetuses from 0.76 to 0.87 gestation, suggesting that maturation of vagal FHR control mainly occurs in NW fetuses. The absent increase of RMSSD and HF band spectral power values in LW fetuses at 0.87 versus 0.76 gestation suggests alteration of maturation of vagal cardiac control in these fetuses. In contrast to the development of vagal FHR control, we could not detect a relationship between fetal body weight or gestational age and sympathetic modulation of FHR. Cardiac sympathetic activity is probably masked by more dominant changes of vagal FHR modulation. This was suggested by Yu & Lumbers (2000) who demonstrated that fHRV is not modulated by baroreceptor-mediated changes in cardiac sympathetic tone in late-gestation fetal sheep. The assumption of masked cardiac sympathetic activity is also supported by the increased sympathetic blood pressure control in LW fetuses shown in the present study. The different trajectory of maturation of the autonomic control of cardiovascular and cardiac function in LW fetuses, rather than an altered vascular reactivity or increased cardiac contractility, is at least a major contributor to the FBP increase

Effects of twin pregnancies In contrast to the NW fetuses, the maturational increase of RMSSD and HF band spectral power as measures of vagal cardiac control was not found in singletons and twins. This might be due to the lower number of fetuses, including some LW fetuses, in these subgroups. At least in singletons, there was a maturational increase of HRVtri, a measure of both sympathetic and vagal modulation of FHR. Otherwise, there were no differences in the maturation of autonomic cardiac control between singletons and twins. Similar to the group of all fetuses, vagal modulation of FHR depended on fetal weight in singletons at 0.87 gestation. The absent relation in twins may reflect delayed vagal maturation as in LW fetuses, but we cannot exclude that it is artificially caused by the low number of fetuses and/or a small variation of weights.

Taken together, the only discrete differences between singletons and twins in the maturation of the FBP and its autonomic control, as well as of the autonomic control of cardiac function, show the minor aetiologic role of twin pregnancies in the different trajectory of development of autonomic control of cardiovascular and cardiac function in LW fetuses. The relatively low effect of twin pregnancies on fetal weight, cardiovascular regulation and autonomic control may be due to the fact that twins belong to normality in sheep. The sheep we studied were taken from an outbred flock. Thus any genetic effects were evenly distributed among singletons and twins. The twin rate is 1.49 in our flock, observed over more than 250 pregnancies. Moreover, we chose the larger of both twins to avoid growth-retarded fetuses. But this means we cannot exclude that we would have found other results in the smaller twin. Nevertheless, our results are in agreement with human twin studies that show little consistent evidence that birth size in twins is associated with increased cardiovascular morbidity or mortality (McMillen & Robinson, 2005). Even studies that examined whether being the smaller rather than the larger twin is consistent with the fetal programming hypothesis did not show consistent results (McMillen & Robinson, 2005).

Conclusion

Our results support the hypothesis that low fetal weight within the normal weight range is associated with altered development of cardiovascular control depending on gestational age. At the beginning of the last third of gestation, i.e. when the cardiovascular system, its autonomic control and the HPA axis are immature, low fetal weight is associated with higher FBP, and baroreceptor reflex sensitivity is highly likely, due to sympathetic activation. After maturation of vagal activity, this association disappears. It is an exciting idea that a FBP increase based on sympathetic activation by adverse environmental influences only occurs before (cardiac) vagal control begins to mature. Maturation of vagal control of cardiac function itself depends on fetal weight, and is altered in conjunction with low fetal weight within the normal weight range. Alteration of autonomic control of cardiovascular and cardiac function was independent of twin pregnancies to a great extent, but twins differed slightly from singletons in the development of autonomic control of cardiac function. Important questions arise from these results. Is maturation of vagal activity causal in preventing sympathetic activation leading to dysfunction of cardiovascular regulation? Are the alterations in vagal cardiac control in the LW fetuses permanent, or do they reflect a delay of development? Answering these questions would help to identify the phases of cardiovascular development that are sensitive to adverse environmental influences and explain the varying results of previous studies. Better knowledge of the effects of the fetal environment on the development of the cardiovascular system and its autonomic control will contribute to better understanding of the mechanisms of fetal programming of susceptibility to cardiovascular diseases in later life. Detection of alterations of the development of autonomic function may help to identify individuals with an increased cardiovascular risk profile in spite of birth weight in the normal weight span.

Ancillary