Introduction
A deficient intake of energy and nutrients during pregnancy that may be associated with intrauterine growth restriction (IUGR) and low birth weight (LBW)Reference Vorster and Kruger 1 , Reference Bergmann, Bergmann, Dudenhausen, Barker, Bergmann and Ogra 2 is an important health problem in underdeveloped countries. Studies in humans and animal models have demonstrated that IUGR, LBW and/or accelerated postnatal weight gain (frequently, but not exclusively associated with IUGR and LBW) could result in an increased prevalence of cardiovascular risk factors and augmented cardiovascular risk.Reference Vorster and Kruger 1 – Reference Gardner, Pearce and Dandrea 8 On the other hand, if nutritional deficiency is limited to an early stage of pregnancy, as frequently occurs in humans due to the intervention of social support programs, it could not be associated with alterations in birth weight or in post-natal growth.Reference Symonds, Stephenson, Gardner and Budge 7 , Reference Schlabritz-Loutsevitch, Ballesteros, Dudley, Jenkins, Hubbard, Burton and Nathanielsz 9 – 11 Data on the cardiovascular impact of nutritional deficiency during pregnancy without IUGR, LBW or accelerated weight gain are scarce and many issues await to be assessed.
Available information regarding the impact of nutritional restriction on the cardiovascular system of offspring shows limitations and controversies. These could be partially explained by methodological factors. First, most studies have used in-vitro techniques to analyze vascular tissues collected from foetuses, new-borns or young adults (i.e., rat or sheep isolated vessels).Reference Dodson, Rozance and Fleenor 3 – Reference Torrens, Brawley and Barker 5 , Reference Cleal, Poore and Boullin 12 – Reference Nishina, Green and McGarrigle 14 However, the functional impact of undernutrition on the cardiovascular system can only be accurately assessed by evaluating the system under real hemodynamic conditions. Second, general studies have focused on a single property, in a specific histological (i.e., muscular)Reference Nishina, Green and McGarrigle 14 or functional (i.e., resistance, conductance) arterial type, but undernutrition could have dissimilar effects on different arteries and arterial properties.Reference Dodson, Rozance, Petrash, Hunter and Ferguson 4 , Reference Torrens, Brawley and Barker 5 , Reference Vianna, Horta, Gigante and de Barros 15 Therefore, results obtained in a certain artery should not be extrapolated to other arteries or arterial pathways. Similarly, alterations in the arterial system may not necessarily affect left ventricle (LV) function (and vice versa).Reference Vlachopoulos, O’Rourke and Nichols 16 Third, available data were mostly obtained from foetal or neonatal tissues, and then, the potential impact of undernutrition in adulthood was theoretically analyzed.Reference Nishina, Green and McGarrigle 14 However, some consequences of intrauterine undernutrition could be observed only in adult life.Reference Armitage, Khan, Taylor, Nathanielsz and Poston 17 Finally, while male offspring have shown to be highly susceptible to intrauterine nutritional restriction, the impact of such restriction on female offspring is not completely accepted. Although in some studies nutritional restriction affected the cardiovascular system of female offspring, it did not in others (i.e., females did not develop hypertension, or if they did, it was not as frequent as in males).Reference Sathishkumar, Elkins, Yallampalli and Yallampalli 18 Further studies are necessary to determine whether the cardiovascular system of female offspring is affected by nutritional restriction during pregnancy. In this context, the studies should consider (1) evaluations in live conscious animals (cardiovascular system working under physiological conditions), (2) the assessment of LV functional capability together with structural and functional properties of different arterial types (i.e., elastic and muscular; conductance and resistance vessels) and (3) the analyses of adult individuals.
We hypothesized that intrauterine nutritional restriction results in impairment of female offspring arterial structure and function observed in adult life. The impact would differ depending on the arterial type, pathways and vascular properties studied, and it would be associated with changes in LV load. In this context, the main objective of this study was to analyze whether nutritional restriction during early and mid-gestation is associated with changes in structural and functional cardiovascular properties in adult female offspring.
Methods
Pregnant ewes (mothers) and female descendants: high and low pasture allowance
This study was performed at the Experimental Station Bernardo-Rosengurtt (32°S, 54°W) from March 2013 to April 2016 (autumn). Thirty-three multiparous Corriedale ewes weighing 46.9 ± 0.9 kg (mean ± standard error of mean [SEM]) and with a body condition score of 2.8 ± 0.1 (1:emaciated to 5:obese)Reference Russel, Doney and Gunn 19 were included and studied in a completely randomized-block design. Ewes grazed on 32 hectares of natural pastures, in three blocks divided into two plots (units) by electric fences. Each treatment was repeated in three plots. Ewes grazed continuously and had free-access to water. From the same cohort of sheep (mothers and descendants), data related to physiological changes and behavioral issues (i.e., ewe–lamb bonding and behaviors at lambing and at weaning) and their association with pasture allowance during pregnancy were previously published.Reference Freitas-de-Melo, Ungerfeld and Hötzel 20 , Reference Freitas-de-Melo, Ungerfeld, Hötzel, Orihuela and Pérez-Clariget 21
From 23 days before insemination until 122 days of gestation, ewes (mothers, “M” prefix) were randomly assigned to one of the two nutritional management methods: (1) normal feeding or high native pasture allowance (M-HPA) or (2) nutritional restriction or low pasture allowance (M-LPA).Reference Freitas-de-Melo, Ungerfeld and Hötzel 20 , Reference Freitas-de-Melo, Ungerfeld, Hötzel, Orihuela and Pérez-Clariget 21 Hence, the “control” group (M-HPA) was constituted by ewes who had normal availability of nutrients during pregnancy, and the under-nourished group of mothers (M-LPA) by ewes exposed to a nutritional restriction (~50–75% of the usual amount of forage offered) during approximately the first 2/3 of pregnancy. Animals from the M-HPA group had access to 10–12 kg/dry matter/100 kg of body weight per day and those from the M-LPA group had access to 5–8 kg dry matter/100 kg of body weight per day. Forage availability was estimated monthly using the double-sampling methodReference Haydock and Shaw 22 and forage allowances were adjusted using “put-and-take” ewes.Reference Freitas-de-Melo, Ungerfeld and Hötzel 20 , Reference Freitas-de-Melo, Ungerfeld, Hötzel, Orihuela and Pérez-Clariget 21 Ewes were shorn on day 122 of gestation, and supplemented with 200 g of rice bran plus 50 ml of crude glycerine (77% of glycerol)/animal/day from the week before shearing until lambing. After shearing, ewes were placed in a paddock with Festuca arundinacea prairie, where they grazed ad libitum. Therefore, before and after the nutritional intervention, M-HPA and M-LPA had access to the same pastures. Body weight and condition score were estimated monthly from 23 days before conception until lambing. Gestational age and birth weight were measured and the birth weight/gestational age ratio was quantified.
All the ewes and their lambs were maintained as a single group for 90 days when lambs were weaned. Then, all female lambs continued grazing together in the same pastures. Sixteen ewes (11 born as single lambs and 5 twins) born from M-HPA (descendants, “D” prefix, or offspring from HPA mothers, “D-HPA” group) and 17 ewes (11 born as singles and 6 twins) born from M-LPA (descendants or offspring from LPA mothers, “D-LPA” group) underwent cardiovascular evaluation when they were 2.5 years old. Body weight, length and condition score were measured at the time of cardiovascular evaluation. The day before cardiovascular evaluation, blood samples were obtained and total cholesterol, glucose, protein and albumin concentrations were determined using an automated chemistry analyzer (WienerLabBT-3000 Plus/CB-350i, Argentina) (Table 1).
Table 1. Characteristics of descendant ewes groups at the time of the cardiovascular study: comparison of high (D-HPA) and low (D-LPA) pasture allowance groups
Values expressed as mean value (MV) and standard error of mean (SE). D-HPA and D-LPA: descendants (offspring) from mother exposed to High and Low Pasture Allowance during pregnancy, respectively. BW: Birth weight. GA: Gestational age. SBP, DBP, PP: systolic, diastolic and pulse blood pressure, respectively. Body condition scoring (BCS): score 1 = emaciated to 5 = obese. A p value < 0.05 was considered significant.
Cardiovascular evaluation of D-HPA and D-LPA ewes
Studies were conducted while the animal was resting quietly in a sling, lying down in dorsal decubitus, being conscious, without anaesthetic administration.Reference Hawkins, Steyn and McGarrigle 23 To avoid isolation (an important stressor in sheep), a companion ewe was continuously present during the evaluation. Evaluation started once the animal was quiet, and hemodynamic variables (i.e., blood pressure [BP], heart rate) were stable. During the cardiovascular studies, researchers did not know the group of the studied ewes belonged to (blinded evaluation).
Evaluation included: (1) peripheral BP (pBP); (2) central (aortic) BP (cBP), aortic wave-derived parameters and LV afterload; (3) common carotid and femoral arteries (CCA and CFA) beat-to-beat diameter waveforms, intima-media thickness and local stiffness; (4) CCA and CFA blood flow velocity levels, patterns and velocity-derived indexes and (5) CCA and CFA characteristic (local) impedance, carotid and femoral pathways (regional) peripheral resistances (Fig. 1a–d). A detailed explanation of the studies and parameters is shown in the Supplementary Methodology.
Fig. 1. (a): arterial non-invasive echographic (B-Mode and Doppler-Mode) and tonometric (applanation tonometry) records. (b, c and d): arterial parameters derived from tonometric or Doppler recordings. SBP, DBP and PP: systolic, diastolic and pulse pressure, respectively. cAP: central augmented pressure. cAIx: central augmentation index. cESP: central end systolic pressure. ED: ejection duration. cSTTI and cDTTI: systolic and diastolic tension time index, respectively. SEVR: subendocardial viability ratio. PSV: peak systolic velocity. EDV: end diastolic velocity. MV: mean velocity. PRV: peak reversal velocity. SPV: secondary forward velocity. MinDV: minimal diastolic velocity (i.e., PRV for CFA). RI: Resistive or Pourcelot Index. PI: Pulsatility or Gosling Index. SDR: Systolic-Diastolic Velocity Ratio or Index.
Peripheral blood pressure
Non-invasive oscillometric pBP measurements (HEM-4030; Omron-Healthcare, USA) were obtained using a cuff placed around the metatarsus of the upper pelvic limb or above the carpus on the upper thoracic limb.Reference Trim, Hofmeister, Peroni and Thoresen 24 Systolic (pSBP) and diastolic (pDBP) pressures were recorded. Then, peripheral pulse (pPP) and mean BP (MBP) pressures were calculated: pPP = pSBP − pDBP and MBP = pDBP + (pPP/3).
Central (aortic) pressure, wave-derived parameters and ventricle afterload
Aortic cBP and wave-derived parameters were assessed (SphygmoCor-CvMS-v.9, AtCor-Medical, Australia) from carotid or central (reference method) and peripheral (femoral) applanation tonometry (CAT and PAT, respectively) (Fig. 1a). First, right CCA BP waveforms were recorded using CAT and calibrated to pDBP and MBP. Then, central systolic, diastolic, end-systolic and pulse pressure (cSBP, cDBP, cESP and cPP) values were obtained from CCA waveforms.Reference Vlachopoulos, O’Rourke and Nichols 16 , Reference Zócalo, Curcio and García-Espinosa 25 Central aortic augmented pressure (cAP) and augmentation index (cAIx = cAP/cPP) were quantified (Fig. 1b).Reference Vlachopoulos, O’Rourke and Nichols 16 LV afterload complementary indexes were calculated: ejection duration, diastolic duration (ms), relative ejection duration (RED = ejection duration/pulse period), ejection duration*cSBP product, RED*cSBP product and subendocardial viability ratio (Fig. 1c).Reference Vlachopoulos, O’Rourke and Nichols 16 Subendocardial viability ratio, an indicator of myocardial perfusion/workload relationship was quantified as the ratio between aortic systolic and diastolic tension-time indexes (cSTTI and cDTTI) (Fig. 1c). cDTTI is the area below diastolic aortic BP curve and cSTTI is the area beneath the systolic aortic BP curve. Right CFA BP waveforms were recorded using PAT.Reference Vlachopoulos, O’Rourke and Nichols 16 , Reference Zócalo, Curcio and García-Espinosa 25 Parameters similar to those obtained from CAT-derived aortic waves were obtained from PAT-derived aortic waves.
Carotid and femoral diameters, intima-media thickness and local stiffness
B-mode ultrasound (7–13 MHz linear-transducer; M-Turbo/SonoSite, USA) was used to obtain sequences of images from longitudinal CCA and CFA views. Beat-to-beat diameter waveforms were obtained using border detection algorithm.Reference Zócalo, Curcio and García-Espinosa 25 , Reference Zócalo, Marotta and García-Espinosa 26 Systolic, mean and end-diastolic diameters and arterial intima-media thickness (computed on the posterior wall at end-diastole) were quantified (Hemodyn-4M/Dinap, Argentina). Pulsate diameter was calculated as systolic diameter minus diastolic diameter. CCA and CFA local stiffness were evaluated using complimentary BP-dependent and independent parameters: pressure–strain elastic modulus, stiffness index, incremental elastic modulus and local pulse wave velocity.Reference Vlachopoulos, O’Rourke and Nichols 16 , Reference Santana, Barra, Grignola, Ginés and Armentano 27
Carotid and femoral blood velocity, flow patterns and blood velocity-derived indexes
Peak systolic, mean, end-diastolic and minimum diastolic velocity levels were computed from CCA and CFA blood flood velocity waveforms (Doppler, 7–13 MHz, M-Turbo-SonoSite, USA) (Fig. 1d). The amplitude and time to early diastolic reversal peak and to the secondary forward diastolic peak were computed in the CFA. Peak systolic, mean and end-diastolic blood flows were determined from blood flow velocities and cross-sectional areas. Blood flow velocity waves were analyzed considering widely used Doppler-derived indexes: resistive Index, pulsatility index and systo-diastolic velocity ratio.Reference Zócalo, Curcio and García-Espinosa 25 , Reference Zócalo, Marotta and García-Espinosa 26 , Reference Bude and Rubin 28
Local characteristic impedance and regional peripheral vascular resistances
From pressure and flow signals,Reference Santana, Barra, Grignola, Ginés and Armentano 27 , Reference Li 29 characteristic impedance was quantified as the ratio between BP (dP/dt; mmHg/s) and blood flow (dF/dt; ml/s) changes observed early in the systolic phase (early ejection). Additionally, carotid and femoral pathways (regional) resistances were quantified as the ratio between mean BP and blood flow.
Data and statistical analysis
After confirming the normal distribution of the variables with the Kolmogorov–Smirnov test, the statistical analysis was divided into three steps. First, concordance between cBP values and aortic wave-derived parameters, obtained with CAT (reference method) and PAT (alternative method), was evaluated in data collected from a subgroup of 12 ewes that showed high-quality CAT recordings and excellent PAT signals. Concordance analysis was done to evaluate the capability of obtaining accurate cBP and wave-derived parameters estimation from peripheral data, since in some animals CAT would not provide reliable results (mainly due to the thickness of the ewes’ neck and the depth of the CCA). Correlation and Bland–Altman analyses were done (Supplementary Table S1 and Figure S1). Positive correlations were observed between methods (CAT and PAT) when cSBP, cDBP, cPP, cESP, subendocardial viability ratio, cSTTI and cDTTI indexes were considered (p < 0.05) (Supplementary Table S1). In turn, no significant correlations were observed for cAIx and cAP (Supplementary Table S1). The systematic difference (mean error) and the slope of the regression equation (proportional error) were not different from zero when cSBP, cDBP, cPP, cESP, cSTTI and cDTTI were considered (Supplementary Table S1 and Figure S1). For subendocardial viability ratio, systematic, but not proportional differences were observed. Then, central BP levels and waveforms obtained using PAT allowed arriving to reliable central parameters (except for cAP and cAIx). Considering those results and given the technical advantages, central hemodynamic and wave-derived parameters were estimated from PAT-derived aortic waveform (Tables 2 and 3).
Table 2. Common carotid and femoral artery blood flow velocities, diameters, wall thickness, local and regional impedance characteristics: comparison of high and low pasture allowance groups
Values expressed as mean value (MV) and standard error of mean (SE). D-HPA and D-LPA: descendants (offspring) from mother exposed to high and low pasture allowance during pregnancy, respectively. a: Adjusted p value: based on estimated marginal means. Covariates appearing in the model are evaluated at the following values: Body weight = 44.609 Kg., Body length = 126.48 cm., Body Condition Scoring (BCS) = 3.48. Zc: characteristic impedance. PVR: peripheral vascular resistances. CI: confidence interval. A p value <0.05 was considered significant.
Table 3. Common carotid and femoral artery stiffness, central (aortic) blood pressure levels and wave-derived left ventricle afterload parameters: comparison of high and low pasture allowance groups
Values expressed as mean value (MV) and standard error of mean (SE). D-HPA and D-LPA: descendants (offspring) from mother exposed to high and low pasture allowance during pregnancy, respectively. EM: elastic modulus. Einc: Incremental elastic modulus. SEVR: subendocardial viability ratio. STTI and DTTI: systolic and diastolic tension time index, respectively. SBP, DBP and PP: systolic, diastolic and pulse blood pressure, respectively. a: Adjusted p value: based on estimated marginal means. Covariates appearing in the model are evaluated at the following values: Body weight = 44.609 kg, body length = 126.48 cm, Body Condition Scoring (BCS) = 3.48. CI: confidence interval. A p value <0.05 was considered significant.
Second, to compare treatments (D-LPA vs. D-HPA) and following the mentioned strategy (Randomized Complete Block Design), the block number was included into the model as a randomized factor. Additionally, we evaluated potential nutritional co-factors (fixed factors) that should be considered for an adequate analysis. In this sense, nutritional characteristics at the time of birth (weight, gestation age, weight/gestational age) or at the time of the cardiovascular study (body weight, length, condition score) were considered (Correlation) (Supplementary Tables S2 and S3).
Third, after analyzing data without considering cofactors, comparisons were made adjusting for body weight, length and condition score (fixed factors previously identified) (Tables 2 and 3).
Taking into account available data (mean values and standard deviations), a total of 32 ewes were the minimum required to detect a statistically significant effect of the different pasture allowances with at least 80% of power (Supplementary Methodology Table S4).Reference Eng 30 Hence, the study sample size (n = 33) was large enough to detect differences maintaining the criteria of reducing the number of experimental units. In any case, comparisons were statistically significant, indicating that the comparisons had adequate statistical power.
Analyses were carried out using MedCalc Statistical Software (Belgium). Statistical differences were considered significant when p < 0.05.
Results
M-HPA had greater body condition score than M-LPA (2.72 ± 0.03 vs. 2.57 ± 0.04; p = 0.004). The number of offspring born from each mother (single or twin), did not impact on the descendants body weight or condition score (weight: 42.9 ± 1.1 vs. 42.5 ± 1.2 kg; body condition score: 3.48 ± 0.07 vs. 3.60 ± 0.06, twins and singles, respectively). There was no association between treatment and number of offspring born from each mother (litter size). There were no differences in birth weight, gestational age or birth weight/gestational age ratio between D-HPA and D-LPA (Table 1). At the time of the cardiovascular study, body weight, length, condition score, hemodynamic and blood variables from D-HPA and D-LPA ewes did not show significant differences (Table 1).
To compare cardiovascular data from D-HPA and D-LPA groups, considering potential co-factors related to nutritional characteristics at birth or at the time of the cardiovascular study, correlation analyses between nutritional and cardiovascular variables were done (Supplementary Data Tables S2 and S3). CFA flow velocities, Doppler indexes (pulsatility index), arterial diameters and intima-media thickness were positively associated with body weight, length and/or condition score. Blood flow velocities, stiffness (stiffness index), diameters and CCA intima-media thickness were also associated with nutritional characteristics at the time of the cardiovascular study. However, except for carotid pathway resistance, cardiovascular parameters were not associated with birth weight, gestational age or birth weight/gestational age ratio. As a result, the comparative analysis between D-HPA and D-LPA groups required considering nutritional variables at the time of the study (but not at birth).
Table 2 shows comparison (D-HPA vs. D-LPA) before and after adjusting for co-factors. CCA end-diastolic velocity was lower in the D-LPA group after considering nutritional characteristics (p = 0.029). In agreement with that, Doppler indexes that evaluate cerebral vascular resistance (resistive index and systo-diastolic velocity ratio) were higher in the D-LPA group (p = 0.030 and 0.022, respectively). Mean pressure/flow ratio data further reinforced those results, since ewes from the D-LPA group showed lower mean (p = 0.042) and end-diastolic (p = 0.028) flows, and a higher resistance to blood flow in the carotid pathway (mainly after adjusting for nutritional characteristics) (p = 0.016). There were no significant differences in CCA characteristic impedance (mainly determined by its cross-sectional area and wall stiffness). Therefore, the higher resistances would be determined by the intracranial carotid territory (cerebral circulation).
The differences in the femoral pathway between D-LPA and D-HPA ewes were not as significant as those described for the carotid pathway. After adjusting for nutritional factors, D-LPA ewes had lower diastolic and mean diameters (p = 0.034 and 0.039, respectively). Despite the lower femoral diameters observed in the D-LPA ewes, CFA characteristic impedance was not lower in that group (p = 0.089).
Table 3 shows that cSTTI and the relative ejection duration*cSBP product (an afterload index) values were higher in ewes from the D-LPA group (p = 0.038 and 0.034, respectively, unadjusted comparison). These results (higher cSTTI and RED*cSBP product) indicate increased LV afterload in D-LPA ewes. It should be noted that when nutritional characteristics were included as cofactors, the differences were statistically significant for cSTTI and RED*cSBP product only with a one-tailed test, suggesting that increased load condition was influenced by the nutritional characteristics of the ewes at the time of the cardiovascular study. The time to and the amplitude of the early diastolic reverse peak and secondary forward peak did not show differences between groups (Table 2).
Discussion
Our work adds support to previous findings mainly obtained in in-vitro studies, as it provides original, complementary information related to the impact of nutritional conditions during pregnancy on offspring cardiovascular parameters. From non-invasive, in-vivo studies in conscious animals, we showed for the first time that ewes that experienced nutritional restriction (50%–75% of control intake) until day 122 of gestation (≈85% term) gave birth to female offspring that at the age of 30 months (2.5 years old) exhibited structural and functional cardiovascular alterations compared to control (D-HPA) ewes. First, D-LPA ewes had higher peripheral vascular resistances and Doppler-derived indexes in the carotid pathway (cerebral). In turn, when structural parameters were analyzed, the impact of nutritional factors was observed on muscular (resistance arteries, CFA) rather than on elastic (conductance arteries, CCA) arteries (Table 2). Second, the vascular changes observed in D-LPA ewes were accompanied by detrimental changes in LV afterload (Table 3). Third, the detrimental vascular changes observed in D-LPA ewes were not associated with birth weight and/or nutritional conditions at the time of evaluation.
Our results add to previous findings from in vitro studies. Increased vascular tone, reduced vasodilator capacity and impaired endothelium-dependent vascular responses (conditions that could result in increased vascular resistance) have been observed in association with nutritional interventions (i.e., rats under a low-protein diet during pregnancy).Reference Franco, Arruda and Dantas 13 , Reference Nishina, Green and McGarrigle 14 , Reference Brawley, Poston and Hanson 31 – Reference Ozaki, Hawkins and Nishina 33 In this study, we did not evaluate endothelium-dependent or -independent vascular responses, but we analyzed functional parameters associated with them in conscious ewes for the first time. In this regard, D-LPA ewes had increased peripheral resistances in the carotid pathway, which are usually associated with endothelial-dependent and independent capability to maintain an adequate basal vasodilator tone (“dilated” microcirculation). It is to note that the increase in regional resistances was observed considering two independent and complementary approaches: blood flow velocity-derived indexes and the relationship between mean BP and flow (Table 2). Resistive index is an indicator of “peripheral” resistances, almost independent of large arteries resistances. In turn, CCA characteristic impedance (“local CCA resistance”) mainly depends on local arterial stiffness (Table 3). Looking at our findings, the increased vascular resistances in D-LPA ewes may be explained by resistive factors associated with the microcirculation and small peripheral arteries.
Carotid pathway resistances were higher in D-LPA than in D-HPA ewes, while there were no differences in the femoral resistances between groups. On the other hand, nutritional restriction was associated with changes in structural parameters (diameter) only when muscular resistance arteries (CFA) were considered, but not when elastic conductance arteries (CCA) were analyzed. Then, our work supports the concept that nutritional restriction during pregnancy could impact differently on offspring conductance and resistance arteries.Reference Dodson, Rozance and Fleenor 3
There were no regional differences in the vascular impact of nutritional restriction during pregnancy when arterial (CCA and CFA) stiffness was analyzed. Furthermore, disregarding the parameter considered (pressure dependent or independent), there were no differences in arterial stiffness between D-LPA and D-HPA ewes. This means that there were no differences in arterial stiffness when considering: (a) potential differences in pBP and cBP (stiffness index), (b) the arterial wall intrinsic stiffness (incremental elastic modulus) and (c) the arterial segment as a three-dimensional structure (pulse wave velocity). At least in theory, the lack of changes in arterial stiffness associated with nutritional restriction could be explained by the increase in nutrient availability during late pregnancy, when structural and functional properties of medium and/or large arteries are established. The extracellular matrix, an arterial stiffness determinant (mainly, elastin and collagen), is formed during the late gestation and the early postnatal period.Reference Dodson, Rozance and Fleenor 3 Smooth muscle cells modulate the extracellular matrix development to achieve the biomechanical requirements of systemic arteries, signalling changes associated with the hemodynamic conditions (i.e., BP and flow within a vessel).Reference Faury 34 Elastin, an extracellular matrix component associated with the arterial stiffness at low BP working conditions, rapidly accumulates during late gestation and the early neonatal period, and is later slowly degraded with aging. Collagen, a stiff component of the extracellular matrix, accumulates in association with increases in biomechanical load (i.e., high BP), aging and disease.Reference Faury 34
The vascular changes observed in D-LPA ewes were accompanied by detrimental changes in LV afterload indexes. In previous studies, LV load was indirectly evaluated by pBP levels (measured on a limb). However, from both, physiological and pathological points of view, aortic BP is the real direct pressure load imposed on the LV during early systole. Consequently, LV structural and functional properties are more associated with cBP than with pBP.Reference Vlachopoulos, O’Rourke and Nichols 16 , Reference Peluso, García-Espinosa and Curcio 35 As a result, an accurate evaluation of LV load would require considering cBP together with the ventricle frequency (heart rate).
It has been demonstrated that experimental undernutrition of pregnant animals is associated with pBP increase in the offspring.Reference Nishina, Green and McGarrigle 14 , Reference Brawley, Poston and Hanson 31 , Reference Barker, Grace and Vivienne 36 , Reference Hanson, Hawkins, Ozaki, O’Brien, Wheeler and Barker 37 In humans, children of mothers who had thin triceps skinfolds in early pregnancy and low weight-gain during pregnancy have increased pBP. In turn, pBP levels in middle-aged men and women are associated with maternal carbohydrate and protein intake during pregnancy.Reference Barker, Grace and Vivienne 36 In this context, our study showed (for the first time) that D-LPA ewes had higher cSTTI and RED*cSBP levels than D-HPA ewes, which means that during “systolic time” BP levels developed by the LV were higher (to overcome aortic pressure), and myocardial oxygen consumption during ejection work was greater in D-LPA ewes. In this sense, the greater the cSTTI and/or RED*cSBP product, the higher the LV load.Reference Vlachopoulos, O’Rourke and Nichols 16 The hemodynamic and vascular characteristics observed in the D-LPA ewes are in agreement with a condition of increased LV load and impaired LV-arterial coupling. Our findings agree with and complement results reported by Cleal et al. Reference Cleal, Poore and Boullin 12 These authors reported that intra-uterine nutritional restriction resulted in increased interventricular septum and mean LV wall thickness in 2.5-year-old sheep. Both our functional findings and the structural results showed by Cleal et al. Reference Cleal, Poore and Boullin 12 suggest that LV load is increased in D-LPA ewes. D-LPA and D-HPA ewes did not show differences in the arrival time or in the amplitude of the early diastolic reverse peak and secondary forward peak. Therefore, the increased LV afterload in D-LPA ewes would not be explained by enhanced reflections from the posterior hemi-body.
Finally, it is noteworthy that the detrimental changes observed in the arterial system of the D-LPA offspring were not strictly associated with the birth weight and/or the nutritional conditions at the time of the cardiovascular evaluation. Our results agree with available data showing that different foetal growth patterns could result in a similar birth size,Reference Dodson, Rozance and Fleenor 3 highlighting the limitations of birth weight as a single reliable indicator of the intrauterine nutritional experience.Reference Barker, Grace and Vivienne 36 Moreover, it has been suggested that nutritional deficiency could have permanent detrimental effects if it occurs in a sensitive period during the intrauterine development. Furthermore, some effects could only be observed late in life.Reference Barker, Grace and Vivienne 36 It can be proposed that placental or foetal compensatory mechanisms that develop in response to maternal nutritional restriction could preserve foetal growth (and hence birth weight), but the consequences of nutritional deficiency during pregnancy may be evident later. Thus, the adoption of a biological “strategy” that ensures intrauterine growth could not ensure well-being in adulthood.Reference Hoet and Hanson 39 In this context, although D-LPA ewes did not show external evidence (anthropometric and nutritional parameters) of the exposure to undernourishment, they had cardiovascular alterations associated with intrauterine nutritional restriction.
Methodological considerations: strengths and limitations
In order to make an adequate interpretation and evaluation of our findings and to analyze strengths and limitations of our experimental approach, some methodological aspects should be considered. There are several factors that make the use of a conscious sheep model a strength. First, there are multiple similarities between human and sheep pregnancy, especially regarding placental development, metabolic function and nutrient transport, making ovine models useful.Reference Symonds, Stephenson, Gardner and Budge 7 , Reference Barry and Anthony 40 Second, ovine and human systemic arteries working conditions and responses to vasoactive agents are quite similar. In this regard, it is noteworthy that our group has experience in evaluating ovine and human cardiovascular systems in-vivousing both, invasive and non-invasive approaches.Reference Zócalo, Curcio and García-Espinosa 25 – Reference Santana, Barra, Grignola, Ginés and Armentano 27 , Reference Bia, Zócalo and Armentano 41 , Reference Bia, Zócalo, Wray and Cabrera-Fischer 42 Third, sheep are practically non-exposed to modifiable cardiovascular risk factors (unlike other animals who consume pro-atherogenic diets), which could modify the association between nutritional interventions and cardiovascular parameters. Then, potential confusing factors that could explain our cardiovascular findings were reduced. Furthermore, to increase the strength of our results, we also made statistical adjustments for potential confusing variables. Taking all these into account, it should be noted that the applicability of our findings to human arteries (and pregnancies) requires confirmation.
It should also be considered that the impact of nutritional deficiency on the arterial system could differ depending on the offspring sex.Reference Sathishkumar, Elkins, Yallampalli and Yallampalli 18 , Reference Dasinger and Alexander 43 In fact, it is currently well accepted that males are highly susceptible to intrauterine nutritional restriction. The inclusion of only females made it possible to work with a homogeneous population, decreasing intra-treatment factors of variation. However, we are aware that it could also represent a limitation, and undoubtedly, analyzing and comparing data from both, males and females, would have been interesting, enriching the work and increasing our knowledge of the impact of nutritional interventions during pregnancy on the cardiovascular system of the offspring. Considering the available information, differences may be even greater in males. However, results cannot be directly extrapolated to male sheep.
In our work, animals were exposed to nutritional restriction only during early and mid-gestation. The model aimed to simulate a real situation mainly observed in humans in developing countries where low accessibility to food in the first two thirds of pregnancy frequently occurs. As pregnancy progresses (second half or final third), many mothers are actively assisted (recruited) by social support programs. Fall described that nutritional interventions in under-nourished pregnant women usually start in the second or third trimester.Reference Fall 44 Therefore, interventions frequently result in newborns without evidence of IUGR or LBW, despite their exposure to intrauterine nutritional restrictions.Reference Fall 44 It is widely known, from the developmental point of view that during the first and second trimesters, the mother is in an anabolic stage (i.e., pregnancy-associated fat accumulation), in the last trimester, she changes to a catabolic stage in which the foetal growth is maximum.Reference Bia, Zócalo, Wray and Cabrera-Fischer 42 Furthermore, during mid-to-late gestation, the foetus undergoes rapid growth.Reference Limones, Sevillano, Sánchez-Alonso, Herrera and Ramos-Álvarez 45 Although, at least in theory, re-feeding or supplementation during the last stage of pregnancy ensures an adequate body growth (weight and height), alterations in organs and vital systems (i.e., cardiovascular system) could persist or develop later. Even though there are many studies on the impact of intrauterine malnutrition on offspring (children) with IUGR and/or LBW, works assessing the impact of malnutrition on descendants who do not show these conditions are scarce. In this context, our study contributes to understanding the impact that nutritional restriction during the first two thirds of pregnancy has on the cardiovascular system of offspring who did not show IUGR or LBW. Other intervention models (i.e., different nutritional restriction severity or patterns and/or interventions during a different period) could associate different alterations. Previous studies showed that foetal responses to changes in maternal nutrition may have immediate benefit for the foetus, but in the long-term it could be detrimental if the postnatal nutritional offer does not match “that predicted by the foetus” on the basis of its prenatal environment. Cleal et al. Reference Cleal, Poore and Boullin 12 observed that the mismatch between pre- and post-natal nutritional environments was associated with impaired cardiovascular function in adult sheep, which was not observed when these environments results were similar (“predictive adaptive responses” hypothesis). In our model, the mismatch between pre- and post-natal nutritional environments was reduced (similar nutrient availability for both groups late in pregnancy and in the post-natal period).
The carotid artery pressure waveform obtained from applanation tonometry is used as a surrogate for the aortic pressure waveform. High-quality pressure waveforms are often easier to obtain in peripheral arteries than in carotid arteries. The use of a specific generalized transfer function allows the aortic pressure waveform to be derived from the peripheral artery pressure waveform. However, the transfer function used by the SphygmoCor system (the device used in this study) was developed and validated in human studies and, to the best of our knowledge, it was not known whether a generalized transfer function could be applicable to the ovine vasculature. Considering the difficulties in obtaining high-quality carotid registers using tonometry, as a first step in our analysis we evaluated whether peripheral registers would allow equivalent levels of central pressure and parameters derived from pulse waves to be obtained. Then, considering the results of the equivalence analysis and given the technical advantages, central hemodynamic and wave-derived parameters were estimated from PAT-derived aortic waveform.
Finally, a joint analysis of our findings showed that an adequate evaluation of the impact of intrauterine nutritional restriction and its impact on the offspring cardiovascular system requires a comprehensive and multiparametric assessment of cardiac and vascular properties, using different techniques and methodological approaches. Different arteries and arterial pathways, as well as both central and peripheral hemodynamic parameters should be considered. If this is not the case, the association between intra-uterine malnutrition and cardiovascular alterations could be under- or overestimated.
Conclusions
Structural and functional cardiovascular parameters were non-invasively assessed in-vivo (conscious animals) in adult ewes with and without intra-uterine exposure to nutritional restriction. Widely used and validated (gold standard) methodological approaches enabled us to evaluate peripheral and central BP, wave-derived parameters, carotid and femoral arteries diameters, wall thickness and stiffness, blood flow levels, local and regional (peripheral) blood flow resistances and LV afterload. At 30 months of age, female offspring exposed to intrauterine nutritional restriction (without evidence of IUGR or LBW) showed higher carotid pathway arterial resistances (cerebral microcirculatory resistances) and LV afterload than those exposed to control nutritional offer. The impact of intrauterine nutritional restriction varied depending on the artery and/or arterial property considered (structural vs. functional), and would not depend on nutritional conditions at birth or in adult life.
Acknowledgments
The authors would like to thank Ing. Agr. Carlos Mantero (Director of the EEBR) for his support throughout this work, to Tec. Pec. Ignacio Sosa for his support in the management of the animals, Ing. Agr. MSc. María José Abud for her support during the experimental phase of the work and Mr. Federico Brum Bazet for his professional proofreading services.
Financial Support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors. Cardiovascular evaluation was partially supported by extra-budgetary funds generated by CUiiDARTE Centre.
Conflicts of Interest
None.
Ethical Standards
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national guides on the care and use of laboratory animals (Guide for the Care and Use of Laboratory Animals published by the US National Research Council, National Academy Press, Washington, DC, 1996, and Comisi;n Honoraria de Experimentaci;n Animal, Universidad de la República, Uruguay) and has been approved by the institutional committee (Comisi;n de Ética en el Uso de Animales, Facultad de Agronomía, Universidad de la República, Uruguay; Ethics committee approval: CEUA-Fagro. 020300-001929-17/111130-001469-13/111130-001856-13).
Supplementary Material
To view supplementary material for this article, please visit http://dx.doi.org/10.1017/S2040174419000230
