Infants with macrosomia and infants of diabetic mothers have increased carotid artery intima-media thickness in childhood

Incidence of diabetes during pregnancy is increasing worldwide, and intrauterine hyperglycemia exposure may have long-term adverse effects on the cardiovascular health of children. We investigated prospectively the risk of atherosclerosis and carotid intima-media thickness (CIMT) in infants born macrosomic and in infants of diabetic mothers (IDM) at the age of 8–9 years in 2021. A total of 49 infants of diabetic mothers (IDM group) and 13 macrosomic infants (macrosomic group) were included in the study. They were compared with 26 age-matched healthy children with birth weight appropriate for gestational age born to non-diabetic mothers (control group). Anthropometric measurements, atherosclerosis risk factors, and CIMT measurements were performed. There was no significant difference between the groups in terms of age, gender, actual anthropometric measurements, blood pressure measurements, laboratory parameters, or atherosclerosis risk factors. Gestational age was lower in the IDM group (p < 0.001), while birth weight was higher in the macrosomic group (p < 0.001). High-density lipoprotein cholesterol level was lower in the IDM group than the other groups. Duration of exclusive and total breastfeeding was lower in IDM group than in the control group (p < 0.001 for both). Body mass index, skinfold thickness, waist-to-hip ratio, and waist-to-height ratio were higher in those breastfed for less than 6 months in the IDM group. The CIMT values were statistically higher in IDM [0.43 ± 0.047 (0.34–0.60)] and macrosomic [0.40 ± 0.055 (0.33–0.50)] groups than control group [0.34 ± 0.047 (0.26–0.45)]. Conclusion: CIMT values were higher in IDM and macrosomic groups at 8–9 years old age compared to children born with normal birth weight. This indicates intrauterine exposure in both groups. And also, breastfeeding seems very important for IDMs. What is Known: • Intrauterine hyperglycemia exposure has long-term adverse effects on the cardiovascular health of children. • Infants of diabetic mothers have higher carotid artery intima-media thickness at birth. What is New: • Both infants of diabetic mothers and infants with macrosomia have increased carotid artery intima-media thickness at the age of 8-9 years. • Duration of breast feeding is important especially in infants of diabetic mothers as body mass index, skinfold thickness, waist to hip and height ratio were higher in those breastfed less than 6 months. What is Known: • Intrauterine hyperglycemia exposure has long-term adverse effects on the cardiovascular health of children. • Infants of diabetic mothers have higher carotid artery intima-media thickness at birth. What is New: • Both infants of diabetic mothers and infants with macrosomia have increased carotid artery intima-media thickness at the age of 8-9 years. • Duration of breast feeding is important especially in infants of diabetic mothers as body mass index, skinfold thickness, waist to hip and height ratio were higher in those breastfed less than 6 months.


Introduction
Incidence of gestational diabetes mellitus (GDM) is increasing worldwide and many babies are exposed to hyperglycemia in utero [1]. Comparison of siblings before and after the development of maternal DM shows the role of epigenetic transmission on long-term negative effects on the cardiovascular health of the offspring [2]. Fetal "poor metabolic memory" resulting from intrauterine hyperglycemia, hyperinsulinemia, and leptin resistance cause offspring obesity, diabetes mellitus (DM) and metabolic disturbances, endothelial dysfunction, and hypertension causing cardiovascular diseases (CVD) [2]. Alterations in insulin-like growth factor-1 and growth hormone axis are also risk factors for CVDs; both hypersecretion and hyposecretion of growth hormone increase the risk [3]. Hyperinsulinemia in the fetus acts as a growth hormone and results in increased fetal fat tissue, glycogen stores, and body mass [4]. Both macrosomic infants and infants of diabetic mothers (IDM) have an increased risk of obesity in childhood and a high growth trajectory, resulting in an emphasis on such factors as antenatal counseling, breastfeeding, and complementary feeding to reduce the effect of this adverse condition [5][6][7].
Increased carotid artery intima-media thickness (CIMT) is a strong predictor of future CVDs, including coronary heart disease, myocardial infarction, stroke, and peripheral artery disease. Dietary factors, lipid profile characteristics, obesity, diabetes, and high blood pressure (BP) are among the risk factors for CVDs [9]. Macrosomic neonates of healthy and diabetic mothers were reported to have significantly higher abdominal aorta intima-media thickness measurements compared to controls [10,11]. A significant association was established between birth weight and CIMT values also in childhood [7].
To the best of our knowledge, there are few studies dealing with IDMs and macrosomic infants during childhood which evaluate multiple atherosclerosis risk factors and CIMT together [12][13][14]. We hypothesized that CIMT may be higher in IDMs and macrosomic infants during childhood at 8-9 years of age compared to children born to healthy mothers and we examined dietary factors, physical activity, obesity, lipid profiles, insulin resistance, blood pressure (BP), and CIMT measurements in the study.

Study design
The study included children of mothers with type 1 DM, Type 2 DM, and GDM who were born in Cukurova University Hospital in 2011 and 2012 (IDM group), and the children of healthy mothers with a birth weight of > 4000 g (macrosomic group). The "control group" consisted of children born in the same years to healthy, non-diabetic mothers, who were appropriate for gestational age and not macrosomic. The study was approved by Çukurova University Ethics Committee (Date: October 4, 2019, Decision No.: 30). Children were enrolled to the study prospectively in 2021.

Exclusion criteria for the patient and control groups
Children with a gestational age of < 34 weeks, birth weight of < 1500 g, with congenital anomalies, chromosomal anomalies, and a history of birth asphyxia, children with any neurological disorder or mental retardation, children with orthopedic disorders that may affect measurements, children with known chronic kidney disease, congenital heart disease (except for hemodynamically insignificant ventricular septal defect, atrial septal defect, and patent ductus arteriosus), Kawasaki disease, childhood cancers, or chronic inflammatory bowel disease, chronic infections (Hepatitis B, C, HIV, etc.), previously diagnosed dyslipidemia, genetic and metabolic disorders, and children whose parents did not give their consent were not included in the study. Blood samples were gained after 8-12 h of fasting and stored at − 80 °C until the biochemical study.

Early atherosclerosis risk factors
Insulin resistance in children: fasting plasma glucose (FPG) and simultaneous plasma insulin levels were measured. A HOMA-IR value of > 2.22 in girls and > 2.67 in boys was considered insulin resistance [15].
Abdominal obesity: an age-and gender-specific waist circumference in the > 90th percentile or a waist circumferenceto-height ratio of ≥ 0.5 was defined as abdominal obesity [16].
Dyslipidemia: the presence of one or more of the following-plasma triglycerides (TG), low-density lipoprotein (LDL) cholesterol and total cholesterol (TC) levels higher than the reference values, or high-density lipoprotein (HDL) cholesterol lower than the reference values-was considered as dyslipidemia [9].
Early CVD: a family history of coronary artery stent, coronary bypass surgery, or stroke in women before the age of 65, and in men before the age of 55 among first-and second-degree relatives was considered as "early cardiovascular disease." Hypertension: according to the pediatric blood pressure categories for children aged 1-13 years, blood pressure was defined as normal, elevated, or hypertension [18].
Skinfold thickness was measured using the skinfold method at the midpoint of the upper arm using a caliper.
Carotid intima-media thickness was assessed by a single experienced reader (AY) who was blinded to the patients and controls via high-resolution ultrasound (Mindray Resona7, Mindray Bio-Medical Electronics Co., Ltd., China) with an L14-6 s linear 5.1-12.5-MHz probe. CIMT was measured three times at the posterior wall, 1 cm distal to the bifurcation of the carotid artery with the child in the supine position [8]. The children were positioned with a small turn to the contralateral side with the neck in the supine position and the mean of the three measurements was noted.

Statistical analysis
IBM SPSS statistics (Version 20.0. Armonk, NY: IBM Corp.) was used for statistical assessment. Categorical data were expressed as numbers and percentages, and continuous data as mean, deviation, median, and interquartile range. Pearson's Chi-square test was used for the comparison of categorical variables; a Mann-Whitney U test was used to compare quantitative data between two independent groups, and a Kruskal-Wallis H test was used to compare three or more groups. Tamhane's T2 test was used for poc-hoc analysis. Correlations between CIMT and age, lipid profile, BMI percentiles, and BP levels were assessed with Pearson's correlation and regression analyses. p < 0.05 was considered statistically significant.

Results
According to inclusion criteria, there were 73 IDMs and 29 macrosomic infants born to healthy mothers at Cukurova University Hospital in 2011 and 2012. All parents were called, but the parents of 60 IDMs and 21 macrosomic infants could be reached and informed about the study. Parents of 49/60 (81.6%) IDM (IDM group) and 13/21 (61.9%) macrosomic infants (macrosomic group) agreed to participate in the study. All children were evaluated prospectively in 2021.
In the IDM group, five mothers had type 1, 19 had Type 2, and 25 had gestational DM. Of these mothers, 36 (73.5%) were treated with insulin during pregnancy. The control group included 26 healthy children who were born to healthy mothers.
Characteristics of the infants in groups are shown in Table 1. There was a statistical difference in birth weight and gestational age between the groups (p < 0.001 for both) with birth weight being higher in the macrosomic group and gestational age lower in the IDM group. Duration of exclusive breastfeeding was longer in control group than in the IDM group (p < 0.001), while there was no significant difference in exclusive breastfeeding between the macrosomic and control groups or between the macrosomic and IDM groups (Table 1). Atherosclerosis risk factors and family history of early CVD of control, IDM, and macrosomic groups were not different ( Table 2). HDL cholesterol level was significantly lower in the IDM group than in the control group (p = 0.006) and macrosomic groups (p = 0.024), Table 2. CIMT measurements of the children were statistically significantly higher in both IDM and Macrosomic groups than in the control group (p < 0.001 and p = 0.008, respectively), there was no difference between IDM and macrosomic groups (Table 3). A total of 34 (38.6%) of children had a BMI of ≥ 85%. There was no difference in CIMT values of those with a BMI of ≥ 85 and < 85 (p = 0.698). The within-group comparisons also yielded no difference (Table 4). No significant correlation was found in control and macrosomic groups between children's CIMT and BMI, systolic BP, diastolic BP, HDL, LDL, TC, and TG. In the IDM group, there was a weak positive correlation between systolic BP and CIMT (r = 0.310, p = 0.030), and a moderate positive correlation between diastolic BP and CIMT (r = 0.486, p < 0.001) values. No significant correlation was found among the other parameters. Evaluating all children together revealed CIMT to be weakly positively correlated with diastolic BP (r = 0.213, p = 0.048), weakly positively correlated with systolic BP (r = 0.204, p = 0.059), and weakly negatively correlated with HDL (r = − 0.371, p < 0.001). Table 5 compares the CIMT values of infants born to mothers with pregestational DM (type 1 DM and type 2 DM)  and GDM. Maternal HbA1C was found to be significantly higher in the pregestational DM group (6.9 ± 1.3 vs. 6.0 ± 0.7), (p = 0.012), while no significant difference was observed in the other parameters such as gestational age, birth weight, weight, BMI, skinfold thickness, and abdominal obesity. In addition, no statistically significant difference was found in CIMT measurements, Table 5. Table 6 compares the IDM group in terms of total breastfeeding durations of < 6 months and ≥ 6 months. BMI, skinfold thickness, waist circumference, waist-to-hip ratio, and waist-to-height ratio were found to be significantly higher in children who had been breastfed for < 6 months in total. Difference in CIMT measurements was borderline significant (p = 0.057).

Discussion
In the present study, we hypothesized that CIMT may be higher in IDMs and macrosomic infants during childhood and we have detected higher CIMT values in macrosomic-born infants and IDMs at 8-9 years old age supporting our hypothesis. The incidence of diabetes during pregnancy is increasing. According to International Diabetes Federation, it is estimated that 16.7% of live births in 2021 had some form of hyperglycemia during pregnancy, the vast majority (80.3%) of which were GDM [1]. Hyperglycemia during pregnancy is one of the leading causes of maternal and fetal morbidity and mortality, and is associated with short-and long-term outcomes. According to Barker's hypothesis of "fetal origins," fetal environment and endocrine status changes lead to adaptations that can permanently change physiology and metabolism, and these adaptations predispose individuals to CVDs and metabolic and endocrine disorders in adulthood [19]. Studies based on this hypothesis suggest that intrauterine exposure to hyperglycemia causes fetal programming that predisposes the infant to obesity, hypertension, CVDs, metabolic syndrome, and Type 2 DM in the long term [2].
The international, multi-ethnic, and observational Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study, which examined the association of maternal hyperglycemia with adverse perinatal outcomes found birth weight and cord C peptide levels of newborns-whose mothers had hyperglycemia but not DM-exhibited a strong and persistent association with maternal glucose levels [20]. HAPO Follow-up Study, in turn, reported maternal glucose levels during pregnancy to be independently associated with childhood adiposity, including overweight, obesity, skinfold thickness, body fat percentage, and waist circumference in pubertal children aged 10-14 years [12]. Similarly, studies of children aged 5-10 years reported a positive correlation between maternal glucose levels during pregnancy and obesity in children [13,14].
In the present study, comparing IDMs and macrosomic infants with control children at 8-9 years old age, gestational    age was lower in the IDM group, while as a result of inclusion criteria birth weight was higher in the macrosomic group (p < 0.001 for both). IDMs are at increased risk of macrosomia and being large for gestational age, and we attribute the similarity of the mean birth weight to that of the control group, but lower than in the macrosomic group to their lower gestational age at birth, as the gestational age of IDMs was 2 weeks less than the other two groups. IDM group had the lowest gestational age, which can be explained by increased risk of both indicated preterm labor and spontaneous preterm labor in diabetic women [21].
Studies about the effects of maternal obesity and overnutrition on metabolic programming in children during prenatal, perinatal, and postnatal periods show that each of the factors indicating excessive food intake, such as prepregnancy maternal obesity and weight gain during pregnancy [22], macrosomic birth [23], and intrauterine hyperglycemic environment [12,20] are associated with childhood obesity and poor cardiovascular outcomes. Macrosomic infants with nondiabetic mothers were also shown to have higher plasma insulin levels [24] and a larger area of pancreatic β-cell hyperplasia than those born at normal weight to nondiabetic mothers [25]. Elevated glucose levels, dyslipidemia, other metabolic alterations, and chronic inflammation are important factors in the atherogenic process. Expansion of adipose tissue induces increased free fatty acid and lipopolysaccharide levels in the blood which triggers an inflammatory response characterized by infiltration of immune cells leading to pro-inflammatory cytokines and then atherosclerosis [26]. There are a lot of serum biomarkers of atherosclerosis such as highly sensitive CRP, adiponectin, F2-isoprostanes, fibrinogen, ghrelin, D-dimer, resistin, cytokines, adhesion molecules, matrix metalloproteinases, pentraxin, myeloperoxidase, oxidized LDL, lipid hydroperoxides, collagen peptides, and microalbuminuria [27,28]. However, a non-invasive and safe technique for the assessment of subclinical vascular disease is CIMT [8]. CIMT measurements are widely used to assess the risk of subclinical vascular disease in adulthood. Similar to adults, research in children has found that hypertension [29], familial hypercholesterolemia [30], overweight and obesity [31], and type 1 DM [32] lead to an increase in CIMT measurements when compared to the age-matched healthy population. The present study established similar rates of risk factors-family history of early CVD, secondhand smoke exposure, obesity, dyslipidemia, and physical inactivity-for atherosclerosis within all groups, while CIMT was higher in the IDM and macrosomic groups compared to the control group. While it is known that atherosclerosis starts during the intrauterine period in IDMs [11,33], our study found higher CIMT values in both the IDM and macrosomic groups, suggesting that macrosomic birth may also have an effect on metabolic programming as in IDM [23][24][25]34]. Similar to our study, Akcakus et al. [11] showed that both macrosomic newborns of diabetic and nondiabetic mothers had higher abdominal intima-media thickness than the control newborns and the lowest HDL cholesterol level was in the macrosomic infants of diabetic mothers.
Diabetes mellitus is not a homogenous disease process. Type 1, 2, and GDM result in dysglycaemia differently during fetal development. In the present study, we did not detect any difference in CIMT values of children whose mothers had preexisting DM or GDM although HbA1C levels of preexisting DM were higher.
Child and adolescent obesity demonstrates a strong association with early atherosclerosis. Carotid IMT is increased in overweight and obese children [35]. Therefore, we examined CIMT measurements in overweight or obese children in our study. CIMT measurements were not different in overweight and obese children overall and within the individual groups. Based on these findings, it can be concluded that CIMT is affected due rather to intrauterine exposure than the extrauterine exposure.
HAPO follow-up study found that exposure to hyperglycemia in utero was significantly associated with elevated plasma glucose levels and insulin resistance in children, independent of maternal and childhood BMIs and family history of diabetes [36]. In the present study, the rate of children with increased HOMA-IR values and impaired FPG was 16.3% (8 children) in the IDM group and 7.6% (2 children) in the control group. In the macrosomic group, there were no children with insulin resistance and impaired FPG. Although the rate of insulin resistance and impaired FPG was higher in the IDM group, the difference was statistically insignificant ( Table 2). International Diabetes Federation [37] recommends oral glucose testing (OGT] as the diagnosing procedure of choice for screening at-risk children for glucose homeostasis abnormalities. Unfortunately, no OGT test was administered in the present study, and so our results might not have fully reflected the abnormality in the glucose metabolism.
In the present study, the lipid profiles of the groups revealed a statistically significant difference in HDL levels and a borderline significant difference in TC levels. The lowest HDL levels were recorded in the IDM group. Akcakus et al. [11] reported lower HDL and higher TC and LDL levels in macrosomic IDM compared to the control group. We may conclude that exposure starting in the intrauterine period in IDMs continues into childhood. In the study by Tam et al. [38] GDM exposure was found to be associated with high systolic BP and diastolic BP, and low HDL levels in children at the age of 7-10 years, similar to our results. However, when the study re-evaluated children at an average age of 15 years, the BP and lipid profiles of the groups were similar [39]. Although our study identified lower HDL levels in IDM aged 8-9 years, the lipid profile of children should be assessed again in different ages to evaluate the long-term effect of maternal hyperglycemia.
Various studies involving children have shown that newly diagnosed hypertension [29], familial hypercholesterolemia [30], overweight and obesity [31], and type 1 DM [32] increase CIMT. In this study, CIMT was weakly positively correlated with diastolic and systolic BP, and borderline weakly negatively correlated with HDL in all children. Similar to the present study, low HDL cholesterol [40] and high BP [41] were associated with an increase in CIMT in children. We can only detect a correlation in the IDM group; there was a weak positive correlation between systolic BP and a moderate positive correlation between diastolic BP and CIMT.
In the present study, a pairwise comparison revealed a significantly shorter duration of exclusive breastfeeding in the IDM group than in the control group. Only 38.7% of the IDM group was breastfed for 12-24 months. We believe that this may be attributed to the more hospitalization rates of IDMs after birth, causing mother and baby separation and leading to decreased milk production. Another factor may be that parents use more formula due to concern that the baby may have hypoglycemia. Breastfeeding during infancy, however, is very important, and as shown in the present study it seems much more important for IDMs. In low-socioeconomic societies, in particular, the duration of breastfeeding should be longer, as the risk of being overweight is 45% lower in children with maternal GDM breastfed for more than 3 months, compared to those breastfed for less than 3 months at the age of 2-8 years [42]. In 6-13 years old aged children, waist circumference, BMI, and visceral and subcutaneous fat were found to be higher in IDMs breastfed for less than 6 months while no difference was found in anthropometric measurements breastfed for more than 6 months, regardless of whether their mothers were diabetic or not [43]. In the present study, the duration of "exclusive breastfeeding" was very low, with a mean of 3 months in IDMs. Similar to previous studies [42,43], significantly higher BMI, waist circumference, skinfold thickness, waist-to-hip, and waist-to-height ratios were recorded in children breastfed for less than 6 months compared with children breastfed for more than 6 months in the IDM group; also CIMT measurement was borderline significantly higher. These findings indicate the importance of breastfeeding in IDMs.
Limitations of our study include the small sample size, especially in the macrosomic group. Furthermore, the fact that the study was conducted in a single healthcare facility makes generalization difficult. Plasma lipid, insulin, and glucose parameters require at least 8 h of fasting, the results can be misleading if the patient is not properly fasted. Some children may not have complied with the fasting requirement. Although we have evaluated physical inactivity and screen time, we have not evaluated their diet.

Conclusion
We found higher CIMT measurements in the IDM group. This supports "Barker's hypothesis," which states that antenatal exposure to hyperglycemia leads to adaptations that permanently alter the fetal physiology and metabolism, and that these adaptations predispose such children to cardiovascular, metabolic, and endocrine disorders in adulthood. The CIMT measurements, however, were also higher in the macrosomic children born to non-diabetic mothers than in the control group, suggesting that factors other than maternal diabetes may also be involved. The shorter duration of breastfeeding in the IDM group and the higher BMI, skinfold thickness, waist circumference, waist circumference-to-hip ratio, and waist circumference-to-height ratio in children breastfed for less than 6 months suggest that breastfeeding is very important in IDMs. Nutrition is very important both in intrauterine and extrauterine periods, especially in IDMs. There is a need for further studies examining similar variables, involving a larger sample.
Authors' contributions Hacer Yapicioglu and Ceren Seckin: investigation, writing, original draft preparation, and editing; Ahmet Yontem: investigation, ultrasound examinations, and statistical analysis; Dincer Yildizdas: investigation and editing. All authors agree to all aspects of the work and approved the final manuscript.
Funding The study was funded by Cukurova University 2021, TTU-13484.

Declarations
Ethics approval This study was performed in line with the principles of the Declaration of Helsinki. The study was approved by Çukurova University Ethics Committee (Date: October 4, 2019, Decision No.: 30).
Consent to participate Written informed consent was obtained from the parents.