**2. Results**

*2.1. Distribution of Children Descending from Normal Pregnancies into Groups Based on Clinical Examination and Consequent Findings*

Children descending from normal pregnancies were divided into two groups based on the results of examination and clinical findings. Children already dispesarized in the department of pediatric cardiology (*n* = 8) and those ones indicated by the sonographer to have valve problems and heart defects tricuspid valve regurgitation (*n* = 8), mitral valve regurgitation (*n* = 1), pulmonary valve regurgitation (*n* = 2), bicuspid aortic valve regurgitation (*n* = 1), ventricular septum defect (*n* = 1), atrial septum defect (*n* = 1), foramen ovale apertum (*n* = 5), arrhythmia (*n* = 1) constituted particular group together with children confirmed over several visits to have a high BP (*n* = 16) (systolic blood pressure (SBP) and/or diastolic blood pressure (DBP) ≥ 90th percentile evaluated by the Age-Based Pediatric Blood Pressure Reference Charts calculator) and/or high BMI (*n* = 9) (BMI 85th percentile evaluated by the BMI Percentile Calculator for Child and Teens). Overall, this group consisted of 38 children (43.18%). The second group consisted from 50 children with normal anamnesis, normal BP, normal BMI, and normal reference values of echocardiographic measurements.

#### *2.2. Up-Regulation of miR-21-5p in Children Descending from Normal Pregnancies that are Overweight/Obese, Prehypertensive/Hypertensive and/or have Abnormal Echocardiogram Findings*

Since we identified within the group of children descending from normal pregnancies those ones with cardiac findings who were already dispesarized in the department of pediatric cardiology, or those ones who were overweight/obese, had prehypertension/hypertension, and/or abnormal echocardiogram findings, we compared the microRNA profile of this group to that one consisting of children with normal anamnesis, normal BP, normal BMI and normal reference values of echocardiographic measurements. The performance of receivers operating characteristic (ROC) curve analysis revealed that only miR-21-5p differentiated children descending from normal pregnancies in dependence on the presence or absence of postnatal abnormal clinical findings with a sensitivity of 28.95% at a specificity of 90.0% (Figure 1).

**Figure 1.** Up-regulation of miR-21-5p in children descending from normal pregnancies that are overweight/obese, prehypertensive/hypertensive and/or have abnormal echocardiogram findings. GH: gestational hypertension; PE: preeclampsia; FGR: fetal growth restriction; ROC: receivers operating characteristic; AUC: Area under the curve; +LR: likelihood ratio positive; –LR: likelihood ratio negative.

### *2.3. Dysregulation of Cardiovascular/Cerebrovascular Disease Associated microRNAs in Children Descending from Complicated Pregnancies*

MicroRNA gene expression was compared between children descending from normal and complicated pregnancies (GH, PE, and FGR). MicroRNA gene expression was analyzed in relation to both aspects, the current presence of cardiovascular risk factors (overweight/obesity and/or prehypertension/hypertension) and cardiovascular complications (valve problems and heart defects) and the previous occurrence of pregnancy-related complications with respect to clinical signs (mild versus severe preeclampsia), dates of delivery (< and > 34 weeks in case of PE, < and > 32 weeks in case of FGR, respectively). The association between microRNA gene expression and Doppler ultrasonography parameters (pulsatility index (PI) in the umbilical artery, PI in the middle cerebral artery, the cerebroplacental ratio, PI in the uterine artery, PI in the ductus venosus, and the presence of unilateral or bilateral diastolic notch in the uterine artery) was analyzed in the group of complicated pregnancies (PE and/or FGR). Just the results that reached a statistical significance or displayed a trend toward aberrant microRNA expression profile in complicated cases are presented below.

#### *2.4. Multiple microRNAs are Up-Regulated in Children Descending from GH Pregnancies*

The ROC curve analysis revealed a significant up-regulation of miR-1-3p, miR-17-5p, miR-20a-5p, miR-21-5p, miR-23a-3p, miR-26a-5p, miR-29a-3p, miR-126-3p, miR-133a-3p, miR-146a-5p, and miR-181a-5p for children descending from GH pregnancies when the comparison to the controls was performed (Figure 2).

**Figure 2.** *Cont*.

**Figure 2.** *Cont*.

**Figure 2.** *Cont*.

**Figure 2.** Postnatal microRNA expression profile in children descending from GH pregnancies. (**<sup>a</sup>**–**k**) Up-regulation of miR-1-3p, miR-17-5p, miR-20a-5p, miR-21-5p, miR-23a-3p, miR-26a-5p, miR-29a-3p, miR-126-3p, miR-133a-3p, miR-146a-5p, and miR-181a-5p was observed in children descending from GH pregnancies when the comparison to the controls was performed.

The sensitivity at 10.0% false positive rate (FPR) for miR-1-3p (46.3%), miR-17-5p (29.63%), miR-20a-5p (20.37%), miR-21-5p (29.63%), miR-23a-3p (27.78%), miR-26a-5p (16.67%), miR-29a-3p (35.19%), miR-126-3p (29.63%), miR-133a-3p (37.04%), miR-146a-5p (18.52%), and miR-181a-5p (31.48%) was found (Figure 2).

#### *2.5. Up-Regulation of miR-21-5p, miR-23a-3p, miR-26a-5p, miR-103a-3p, miR-125b-5p, miR-195-5p, and miR-342-3p in Children with Normal Postnatal Clinical Findings Descending from GH Pregnancies*

Concurrently, it was observed that the expression of miR-21-5p, miR-23a-3p, miR-26a-5p, miR-103a-3p, miR-125b-5p, miR-195-5p, and miR-342-3p differed significantly between the groups of children affected with GH with normal postnatal clinical findings and the controls (Figure 3). The sensitivity of individual microRNAs at 10.0% FPR was the following: miR-21-5p (39.13%), miR-23a-3p (34.78%), miR-26a-5p (21.74%), miR-103a-3p (30.43%), miR-125b-5p (47.83%), miR-195-5p (34.78%), and miR-342-3p (21.74%) (Figure 3).

**Figure 3.** *Cont*.

**Figure 3.** *Cont*.

**Figure 3.** Postnatal microRNA expression profile in children with normal postnatal clinical findings descending from GH pregnancies. (**<sup>a</sup>**–**g**) Up-regulation of miR-21-5p, miR-23a-3p, miR-26a-5p, miR-103a-3p, miR-125b-5p, miR-195-5p and miR-342-3p was observed in children with normal postnatal clinical findings descending from GH pregnancies.

Screening based on the combination of miR-26a-5p and miR-195-5p showed the highest accuracy for children with normal clinical findings with a prior exposure to GH (AUC 0.717, *p* = 0.001, sensitivity 86.96%, specificity 52.0%, cut off > 0.246824). It was able to identify 34.78% children with an increased cardiovascular risk at 10.0% FPR (Figure 4).

**Figure 4.** Combined postnatal screening of microRNAs in the identification of children with normal postnatal clinical findings descending from GH pregnancies. Postnatal combined screening of miR-26a-5p and miR-195-5p showed the highest accuracy for the identification of children with normal clinical findings with a prior exposure to GH at a higher risk of later development of cardiovascular/cerebrovascular diseases.

#### *2.6. Up-Regulation of miR-20a-5p in Children with Abnormal Postnatal Clinical Findings Descending from GH Pregnancies*

Overall, the group with abnormal postnatal clinical findings consisted of 31/54 children (57.41%) exposed to GH, 5 children already dispesarized in the department of pediatric cardiology, 12 children with abnormal echocardiographic findings (6 tricuspid valve regurgitation, 1 mitral valve regurgitation, 3 pulmonary valve regurgitation, 1 bicuspid aortic valve regurgitation, 1 ventricular septum defect, 2 foramen ovale apertum), 17 children with prehypertension/hypertension, and 3 children with high BMI. miR-20a-5p expression differed between the groups of children affected with GH with abnormal postnatal clinical findings and the controls. miR-20a-5p differentiated between the children descending from pregnancies affected with GH with abnormal postnatal clinical findings and the controls with a sensitivity of 25.81% at 10.0% FPR (Figure 5).

**Figure 5.** Increased expression of miR-20a-5p in children with abnormal postnatal clinical findings descending from GH pregnancies. Increased expression of miR-20a-5p was found in children with abnormal postnatal clinical findings descending from GH pregnancies.

#### *2.7. Up-Regulation of miR-1-3p, miR-17-5p, miR-29a-3p, miR-126-3p, miR-133a-3p, miR-146a-5p, and miR-181a-5p in Children Descending from GH Pregnancies Irrespective of Postnatal Clinical Findings*

The ROC curve analysis showed the difference in gene expression of miR-1-3p, miR-17-5p, miR-29a-3p, miR-126-3p, miR-133a-3p, miR-146a-5p, and miR-181a-5p between the controls and the group of children exposed to GH with postnatal normal clinical findings or children with a prior exposure to GH that already developed any cardiovascular complication (valve problems and heart defects) or were identified to be overweight/obese and/or prehypertensive/hypertensive (Figure 6). The sensitivity at 10.0% FPR for miR-1-3p (47.83% versus 45.16%) and miR-17-5p (30.43% versus 29.03%) for children descending from GH pregnancies with postnatal normal and abnormal clinical findings was approximately equal (Figure 6). The sensitivity at 10.0% FPR for miR-29a-3p (39.13% versus 32.26%), miR-126-3p (39.13% versus 22.58%), miR-133a-3p (43.48% versus 32.26%), miR-146a-5p (26.09% versus 12.9%), and miR-181a-5p (39.13% versus 25.81%) was slightly higher for children descending from GH pregnancies with postnatal normal clinical findings when compared to those ones with postnatal abnormal clinical findings (Figure 6).

**Figure 6.** *Cont*.

**Figure 6.** *Cont*.

**Figure 6.** *Cont*.

**Figure 6.** Postnatal microRNA expression profile in children descending from GH pregnancies irrespective of postnatal clinical findings. (**<sup>a</sup>**–**g**) Increased expression of miR-1-3p, miR-17-5p, miR-29a-3p, miR-126-3p, miR-133a-3p, miR-146a-5p, and miR-181a-5p was observed in children descending from GH pregnancies with normal or abnormal postnatal clinical findings. The ROC curve analysis showed the difference in microRNA gene expression between the controls and the group of children exposed to GH with postnatal normal clinical findings or children with a prior exposure to GH that already developed any cardiovascular complication (valve problems and heart defects) or were identified to be overweight/obese and/or prehypertensive/hypertensive.

Combined screening of miR-1-3p, miR-29a-3p, miR-126-3p, miR-133a-3p, and miR-181a-5p showed the highest accuracy for children with a prior exposure to GH with normal clinical findings (AUC 0.803, *p* < 0.001, sensitivity 82.61%, specificity 74.0%, cut off > 0.224754). It was able to identify 47.83% children with an increased cardiovascular risk at 10.0% FPR (Figure 7).

In addition, the combination of all seven microRNA biomarkers (miR-1-3p, miR-17-5p, miR-29a-3p, miR-126-3p, miR-133a-3p, miR-146a-5p, and miR-181a-5p) may be used to identify children with a prior exposure to GH with abnormal clinical findings (AUC 0.801, *p* < 0.001, sensitivity 70.97%, specificity 76.0%, cut off > 0.353483). It was able to identify 38.71% children with an increased cardiovascular risk at 10.0% FPR (Figure 8).

#### **Figure 7.** Combined postnatal screening of microRNAs in the identification of children with normal postnatal clinical findings descending from GH pregnancies. Postnatal combined screening of miR-1-3p, miR-29a-3p, miR-126-3p, miR-133a-3p and miR-181a-5p showed the highest accuracy for the identification of children with normal clinical findings with a prior exposure to GH at a higher risk of later development of cardiovascular/cerebrovascular diseases.

**Figure 8.** Combined postnatal screening of microRNAs in the identification of children with abnormal postnatal clinical findings descending from GH pregnancies. Postnatal combined screening of miR-1-3p, miR-17-5p, miR-29a-3p, miR-126-3p, mir-133a-3p, miR-146a-5p, and miR-181a-5p showed the highest accuracy for the identification of children with abnormal clinical findings with a prior exposure to GH at an increased risk of later onset of cardiovascular/cerebrovascular diseases.

### *2.8. Cardiovascular/Cerebrovascular Disease Associated microRNAs are Dysregulated in Children Descending from PE Pregnancies*

Overall, the group with abnormal postnatal clinical findings consisted of 63/133 children (47.37%) exposed to PE 19 children already dispesarized in the department of pediatric cardiology, 18 children with abnormal echocardiographic findings (7 tricuspid valve regurgitation, 2 mitral valve regurgitation, 1 pulmonary valve regurgitation, 1 ventricular septum defect, 1 atrial septum defect, 1 hypertrophic ventricular septum, 9 foramen ovale apertum, and 1 arrhythmia), 36 children with prehypertension/hypertension and 5 children with high BMI. Abnormal clinical findings were found in 16/27 children affected with mild PE (59.3%), 47/106 children exposed to severe PE (44.3%), 26/49 children exposed to early PE (53.1%), and 37/84 children affected with late PE (44.0%).

#### *2.9. Increased Expression of miR-133a-3p in Children Descending from PE Pregnancies*

The ROC curve analysis was able to identify up-regulated expression profile of miR-133a-3p in 22.56% children affected with PE regardless of the severity of the disease and the delivery date at 10.0% FPR (Figure 9a).

**Figure 9.** *Cont*.

**Figure 9.** *Cont*.

**Figure 9.** Increased expression of miR-133a-3p in children descending from PE pregnancies. (**<sup>a</sup>**–**<sup>c</sup>**) Increased expression of miR-133a-3p was observed in children descending from PE pregnancies regardless of the severity of the disease and delivery date, severe PE and late PE; (**d**,**<sup>e</sup>**) Increased expression of miR-133a-3p was found in children with both normal and abnormal postnatal clinical findings previously exposed to severe PE and late PE; (**f**) Increased expression of miR-133a-3p was found in children with normal postnatal clinical findings previously exposed to early PE.

Subsequently, miR-133a-3p differentiated between children affected with severe PE and controls with a sensitivity of 23.58% at a specificity of 90.0% (Figure 9b).

Parallel, miR-133a-3p was able to identify children exposed to late PE with a sensitivity of 21.43% at a specificity of 90.0% (Figure 9c).

In addition to that, the consecutive analysis showed that the accuracy of miR-133a-3p biomarker for severe PE affected children identified to have normal or abnormal clinical findings was 25.42% and 21.28% sensitivity at 10.0% FPR (Figure 9d).

Similarly, postnatal screening of miR-133a-3p was able to identify children with a history of late PE with either normal or abnormal clinical findings with a sensitivity of 19.15% and 24.32% at a specificity of 90.0% (Figure 9e).

Parallel, ROC curve analysis of miR-133a-3p identified a significant proportion of children with normal clinical findings exposed to early PE during gestation (a sensitivity of 39.13% at 10.0% FPR) (Figure 9f).

#### *2.10. Increased Expression of miR-1-3p, miR-20a-5p, miR-103a-3p, and miR-342-3p in Children with Abnormal Clinical Findings Descending from PE Pregnancies*

The sensitivity at 10.0% FPR for miR-1-3p for children exposed to late PE was 27.38% and miR-1-3p showed even a higher performance for children with abnormal postnatal clinical findings that were exposed to late PE (a sensitivity of 35.14% at a specificity of 90.0%) (Figure 10a).

Furthermore, at 10.0% FPR, miR-103a-3p was up-regulated in 12.77% and 13.51% children with abnormal clinical findings that were exposed to severe PE or late PE (Figure 10b,c).

In addition, miR-20a-5p was up-regulated in 13.51% children with abnormal clinical findings with a prior exposure to late PE (Figure 10d).

MiR-342-3p represented the unique marker, which was able to differentiate between the group of children with abnormal clinical findings that were exposed to early PE and the controls (a sensitivity of 26.92% at 10.0% FPR) (Figure 11).

**Figure 10.** *Cont*.

**Figure 10.** Postnatal microRNA expression profile in children with abnormal postnatal clinical findings descending from PE pregnancies. (**a**) Increased expression of miR-1-3p was found in children with abnormal postnatal clinical findings exposed to late PE; (**b**,**<sup>c</sup>**) Increased expression of miR-103a-3p was observed in children with abnormal postnatal clinical findings exposed to severe PE or late PE; (**d**) Increased expression of miR-20a-5p was found in children with abnormal postnatal clinical findings with a prior exposure to late PE.

**Figure 11.** Decreased expression of miR-342-3p in children with abnormal postnatal clinical findings descending from early PE pregnancies.

The combination of miR-103a-3p and miR-133a-3p (AUC 0.637, *p* = 0.015, sensitivity 70.21%, specificity 56.0%, cut off > 0.429300) was superior over using only individual microRNA biomarkers, since it was able to identify at 10.0% FPR within the group of children with abnormal clinical findings previously exposed to severe PE 21.28% children with an increased cardiovascular risk (Figure 12).

**Figure 12.** Combined postnatal screening of microRNAs in the identification of children with abnormal postnatal clinical findings descending from severe PE pregnancies. Postnatal combined screening of miR-103a-3p and miR-133a-3p showed the highest accuracy for the identification of children with abnormal clinical findings with a prior exposure to severe PE at a higher risk of later development of cardiovascular/cerebrovascular diseases.

Parallel, combined screening of four microRNAs (miR-1-3p, miR-20a-5p, miR-103a-3p, and miR-133a-3p) showed the highest accuracy for children with abnormal clinical findings with a prior exposure to late PE (AUC 0.701, *p* < 0.001, sensitivity 59.46%, specificity 72.0%, cut off > 0.391116). It was able to identify 32.43% children with an increased cardiovascular risk at 10.0% FPR (Figure 13).

**Figure 13.** Combined postnatal screening of microRNAs in the identification of children with abnormal postnatal clinical findings descending from late PE pregnancies. Postnatal screening based on the combination of miR-1-3p, miR-20a-5p, miR-103a-3p, and miR-133a-3p showed the highest accuracy for the identification of children with abnormal clinical findings with a prior exposure to late PE at a higher risk of later development of cardiovascular/cerebrovascular diseases.

#### *2.11. Up-Regulation of miR-20b-5p in Children with Normal Clinical Findings Descending from Mild PE Pregnancies*

MiR-20b-5p was the only biomarker that differentiated between the group of children with normal clinical findings that were exposed to mild PE and the controls (a sensitivity of 36.36% at 10.0% FPR) (Figure 14).

**Figure 14.** Increased expression of miR-20b-5p in children with normal postnatal clinical findings descending from mild PE pregnancies.

*2.12. Dysregulation of Cardiovascular/Cerebrovascular Disease Associated microRNAs in Children Descending from FGR Pregnancies*

Overall, the group with abnormal postnatal clinical findings consisted of 22/34 children (64.7%) exposed to FGR (9 children already dispesarized in the department of pediatric cardiology, 9 children with abnormal echocardiographic findings (4 tricuspid valve regurgitation, 3 pulmonary valve regurgitation, 1 atrial septum defect, 6 foramen ovale apertum), 9 children with prehypertension/hypertension and 1 child with high BMI). Abnormal clinical findings were found in 7/13 children affected with early FGR (53.85%), and in 15/21 children exposed to late FGR (71.43%).

*2.13. Increased Expression of miR-17-5p, miR-126-3p and miR-133a-3p in Children with Abnormal Clinical Findings Descending from FGR Pregnancies*

MiR-17-5p, miR-126-3p, and miR-133a-3p differentiated between the group of children with abnormal clinical findings that were affected with FGR and the controls with a sensitivity of 22.73 %, 31.82% and 31.82% at a specificity of 90.0% (Figure 15a–c).

The combination of all examined microRNAs (AUC 0.710, *p* = 0.002, sensitivity 63.64%, specificity 78.0%, cut off > 0.305781) was superior over using only individual microRNA biomarkers, since it was able to identify at 10.0% FPR within the group of children with abnormal clinical findings affected with FGR 40.91% children with an increased cardiovascular risk (Figure 16).

**Figure 15.** Postnatal microRNA expression profile in children with abnormal postnatal clinical findings descending from FGR pregnancies. (**<sup>a</sup>**–**<sup>c</sup>**) Increased expression of miR-17-5p, miR-126-3p, and miR-133a-3p was observed in children with abnormal postnatal clinical findings descending from FGR pregnancies.

**Figure 16.** Combined postnatal screening of microRNAs in the identification of children with abnormal postnatal clinical findings descending from FGR pregnancies. Postnatal screening based on the combination of miR-17-5p, miR-126-3p and miR-133a-3p showed the highest accuracy for the identification of children with abnormal clinical findings with a prior exposure to FGR at a higher risk of later development of cardiovascular/cerebrovascular diseases.

#### *2.14. The Association between Postnatal Expression of miR-210-3p and the Severity of PE and/or FGR with regard to Doppler Ultrasonography Parameters*

The index of pulsatility in the ductus venosus (DV) showed a strong positive correlation with miR-210-3p gene expression in children with a history of PE and/or FGR (Figure 17). That means that children with prior findings of DV dilatation indicating poor outcome in severe FGR demonstrated increased postpartum levels of miR-210-3p.

**Figure 17.** Association between postnatal miR-210-3p expression and the pulsatility index in the ductus venosus in PE and/or FGR patients. The pulsatility index in the ductus venosus showed a strong positive correlation with miR-210-3p gene expression in patients with a history of PE and/or FGR.

The results of aberrant expression profile of microRNAs in children descending from pregnancy-related complications are summarized in Table 1.

*Int. J. Mol. Sci.* **2019**, *20*, 654 **Table 1.** Aberrant expression profile of microRNAs in children descending from pregnancy-related complications.


#### *2.15. The Effect of Children Age on Particular microRNA Expression in Children Descending from Normal and Complicated Pregnancies*

The effect of children age on particular microRNA expression in children descending from normal and complicated pregnancies is discussed in supplementary file (Figures S1–S4).
