Epicardial fat thickness in infants of diabetic mothers

1 Introduction

The requirement of intensive care in infants of diabetic mother (IDM) in the first days of life has diminished because of the adequate metabolic control in the mother [1]. However, despite the advances in perinatal medicine, IDM have a high risk for developing perinatal complications, congenital malformations, and in long term, metabolic syndrome [2, 3].

For many years, asymmetrical septal hypertrophy and hypertrophic cardiomyopathy (HCM) have been reported as the result of insulin and others growth factors that stimulate the heart of infants in diabetic mothers [4]. Recently, it has been described that the overcharged of fatty acids at cardiac level, specifically at epicardial fat (EF) specially studied in adults, is an early contributor in the diastolic dysfunction, similar to alterations found in HCM. EF is a visceral and a metabolically active fat depot in anatomical and functional contiguity with the myocardium and the coronary arteries [5, 6]. It produces adipokines and chemokines, which have paracrine and vasocrine effects, playing an important role between the association of obesity and coronary disease [6, 7]. Although, it has been considered as local energy source in high demand situations, EF is also a defense against hypothermia and ischemia, and a secretor of anti-inflammatory molecules like adiponectin and adrenomedullin [8–10]. In pathological conditions such as obesity or diabetes mellitus, it is able to liberate pro-inflammatory and proaterogenic molecules, but the equilibrium between the protector and harmful effects of this adipose tissue is not clear [11]. A higher epicardial fat thickness (EFT) has been found in pregnant women with gestational diabetes and in their fetus than in non-diabetic mothers and their fetus [12]. The objective of this study was to compare the EFT in IDM versus infants of non-diabetic mothers (INDM) and its association with other echocardiographic parameters.

2 Methods

We performed a cross sectional study in 93 infants (64 IDM and 29 INDM) in the General Hospital, Leon, Mexico. Infants with diagnosis of congenital heart disease or respiratory diseases in their medical chart review were excluded. None of these neonates showed metabolic or other acute complications that needed extended hospitalization or re-hospitalization during the first 28 days of life.

3 Results

No difference in sex distribution, weight, and size was observed between groups. However, gestational age was lower in IDM and born more frequently by cesarean than in the control group (Table 1).

Forty-eight (75%) mothers in the diabetic group of neonates showed gestational diabetes, 15 (23.43%) type 2 diabetes and only one (1.56%) type 1 diabetes. During the study, 46 (71.85%) of them were exclusively on diet control, 13 received insulin (20.31%) plus diet, and 5 received insulin and metformin plus diet (7.81%). We identified that 34 (53.12%) mothers in this group showed HbA1C > 6%. Furthermore, the highest value was 9.6%in only one mother as an outlier value, and 10 (15.62%) showed HbA1C≥7%.

EFT, IVST and IVST/LVPW were higher in IDM, while left ventricular fraction ejection (LVFE) was lower than in INDM (Table 2). The ASH frequency in IDM was 45.3 %, and in 3.4%of the control group.

No difference in EFT was found in neonates according to the different type of DM in their mothers. Because of the lack of definite value considered normal for EFT in neonates, we decided to use the highest value in the control group, which was 3.3 mm. According to this, we identified that 10 (66.6%) IDM with Type 2 diabetes, and 28 (58.3%) IDM with gestational diabetes showed high EFT values. In the unique case of IDM with type 1 diabetes, EFT value was exactly 3.3 mm. No correlation was observed in EFT values in neonates and HbA1c in their mothers (r = 0.061; p = 0.937).

Aortic peak velocity and pulmonary integral velocity were lower, while mitral integral early velocity was higher in IDM than in INDM (Table 3).

We found a significant positive correlation between EFT with IVST, LVPW, IVST/LVPW index, and mitral integral early velocity, while a negative correlation with LVFE was observed (Table 4).

EFT has emerged as a new marker of cardio-metabolic risk due to its intimate proximity with the myocardium, the coronary arteries and its endocrine-metabolic features [8]. It is consistently associated with metabolic syndrome (MS), and it varies according to the ethnicity and age showing a value < 7 mm in healthy and slim adult subjects [19]. In children 2.5±0.81 mm for non-obese and 5.5±1.98 mm for obese children have been reported, and EFT > 5 mm seems to show a good sensibility to predict MS in children and adolescents [20, 21]. Furthermore, maternal diabetes upsets the intrauterine environment resulting in an excessive growth, high weight at birth, and ASH [22]. It is possible that maternal diabetes since an early stage changes the dimensions of the EFT, because it has been reported higher in diabetic mother fetuses than in the non-diabetic mother fetuses [12, 23]. In our study, EFT was higher in IDM than in INDM as well as the interventricular septum thickness (IVST) and the ASH frequency, as previously identified in IDM or in children with family history of type 2 diabetes mellitus [24].

We found a positive correlation between EFT with IVST, LVPW, IVST/LVPW index, and mitral integral early velocity, while a negative correlation with LVFE was observed. A positive correlation between EFT with IVST and IVST/LVPW has been previously reported by Eren et al. in obese children [25], pointing secondary to the intimate contact of the EFT with the myocardium [20]. Furthermore, secretory products from EFT from patients with T2D have also been shown to depress cardiomyocyte contractile function and fat oxidation, which could explain the negative relationship with LVFE [26]. Nonetheless, we can not only consider a deleterious effect of EFT in neonates, because cardiac cryoprotection due to its thermogenic capacity, resembling that of brown/beige adipocytes has also been reported [27, 28]. So, this higher EFT in IDM could reflect fetal adaptation to the adverse in utero environment [29].

Although some valvular peak and integral velocities were different between groups, all these values are in normal ranges [30]. Furthermore, changes in peak systolic velocities in neonates must consider the loading condition when interpreting the values and the interpretation of diastolic velocities is more challenging because more factors influence the measurement. EFT threshold values are not identified in neonates and it needs to be clarified whether EFT is a clinically relevant risk factor that will improve risk stratification in this population and guide future clinical decision-making. Nonetheless, considering that several studies have shown that EFT is associated with diastolic dysfunction at older age [31–33], long term studies in this population are desirable, and using E/A ratio, E/e’ ratio, and tissue Doppler echocardiography for diastolic function evaluation.

It has been described that a good prenatal glycemia control, especially in the 3rd trimester can diminish the frequency of neonatal complications, such as macrosomia but not the tendency to cardiac growth and septal hypertrophy [34]. In this study, no association between the maternal diabetes control and the EFT dimensions was found in contrast with fetal EFT in diabetic mothers [35]. As a limitation in our study, only diabetic mothers had HbA1C values, and only eutrophic patients were recruited. We didn’t find any report of hypoglycemia in our patients, the lack of complications is apparently related to the fact we only evaluated eutrophic infants.

Whether an echocardiogram will be performed and include the EFT measure, not enough long term studies can support this recommendation. Even some authors recommend EFT intrauterine monitorization, because it is considered as a novel marker to the fetus exposure to glucose, existing already a nomogram for its intrauterine monitorization [33], and it is considered that this can help to improve the diabetic fetopathy management and diagnosis [25].

The strength in our study lies in the fact that we are the first to report the relationship between EFT with other cardiac parameters in IDM. However, as a limitation of this study, the evaluation of the LV ejection fraction was not performed with modified Simpson method (biplane method), instead of linear measurement. Thus, the present results provide support for more convincing measurements to support our findings and a prospective study to further investigate the association between newborn EFT and adiposity, both in terms of body fat percentage and the metabolic profile and cardiac function in other infant stages of development. Nevertheless, studies including long term following and its change in the medium term and their interactions with factors that have been described as protective such as exclusive breastfeeding, weight control, physical activity, etc., must be carried in this population.

5 Conclusion

EFT was significantly higher in IDM than in INDM and no relationship was found with metabolic control. EFT was positively related with echocardiographic parameters of left ventricular thickness and negatively with left ventricular ejection function.

Disclosure statements

The authors have no conflicts of interest relevant to this article.