Mechanisms linking preterm birth to onset of cardiovascular disease later in adulthood

Abstract

Cardiovascular disease (CVD) rates in adulthood are high in premature infants; unfortunately, the underlying mechanisms are not well defined. In this review, we discuss potential pathways that could lead to CVD in premature babies. Studies show intense oxidant stress and inflammation at tissue levels in these neonates. Alterations in lipid profile, foetal epigenomics, and gut microbiota in these infants may also underlie the development of CVD. Recently, probiotic bacteria, such as the mucin-degrading bacterium Akkermansia muciniphila have been shown to reduce inflammation and prevent heart disease in animal models. All this information might enable scientists and clinicians to target pathways to act early to curtail the adverse effects of prematurity on the cardiovascular system. This could lead to primary and secondary prevention of CVD and improve survival among preterm neonates later in adult life.

Introduction

Preterm birth, defined as live birth before 37 weeks of gestation, is a significant global health problem. According to data provided by the Center for Disease Control and Prevention for 2016, 10% of all babies born in USA were preterm.¹ Prematurity is associated with significant health problems including chronic lung disease, neurodevelopmental delays, periventricular leukomalacia, retinopathy of prematurity, and increased risk of infant death.^2,3 Several studies have been conducted to determine long-term outcomes of preterm infants into adulthood, reporting associations between prematurity and high blood pressure, Type 2 diabetes, and stroke during adult life.^4,5 Various epidemiological studies have suggested that events in utero and early during life can influence body composition and metabolism, thereby increasing the occurrence of a variety of adult diseases, including hypertension, Type 2 diabetes mellitus, and atherosclerotic cardiovascular disease (CVD).^6,7 Barker⁸ first postulated that impaired intrauterine and post-natal growth may be associated with increased risk of CVD in adult life. Furthermore, babies exposed to the famine epidemic in early gestation after the Second World War in the Netherlands had an increased incidence of obesity and CVD later in life.⁹ Following this, many investigators showed a strong association between prematurity and CVD risk factors such as increased systolic and diastolic blood pressures, impaired glucose tolerance, increased insulin resistance, hypertriglyceridaemia, and low high-density lipoprotein levels in serum.¹⁰

Based on a study of 4630 men born in Helsinki University Hospital, Eriksson et al.¹¹ observed that low ponderal index (ratio of body weight to length) along with low weight gain in the 1st year of life had a markedly increased risk of developing ischaemic CVD later in life. These authors reported that premature infants with rapid gain in weight and body mass index in infancy and early childhood (between age 1 year and 12 years) had an increased risk of developing CVD, an effect confined to children with a ponderal index <26 at birth.¹¹ Others have also observed low birthweight as a forewarning for CVD^12,13; however, low birthweight is a crude marker for intrauterine growth and is not necessarily caused by prematurity. Other studies that focused on preterm infants indicated gestational age as a factor in the development of CVD. One such study, conducted in a cohort from 1986 in Northern Finland, showed that preterm girls had higher blood pressure (+6.7 mmHg systolic blood pressure and +3.5 mmHg diastolic blood pressure) in adolescence. This study also showed that boys born preterm in comparison to girls had 6.7% higher total cholesterol, 11.7% higher LDL-C, and 12.3% (3.1–22.4%) higher apolipoprotein B concentrations.¹⁴ These gender-specific differences in adolescents born prematurely persisted even after birthweight adjustment. On average, preterm birth was associated with higher LDL-C levels¹⁵ and elevated systolic and diastolic blood pressure¹⁶ compared to full-term born adults. Twin studies including dizygotic and monozygotic groups showed that along with intrauterine environment influences, genetic factors play a role in development of CVD later in life.¹⁷ Despite the association between prematurity and risk for developing CVD, the mechanistic basis for this association is not well defined. As shown in Take home figure, we review some possible mechanisms by which the observed associations can be explained.

Take home figure.

Mechanisms linking preterm birth and cardiovascular disease in a nutshell.

Potential mechanisms linking preterm birth to cardiovascular disease

Intracellular oxidative stress, diminished nitric oxide generation, and endothelial dysfunction

From the onset of atherosclerosis to culmination into a clinically relevant event, a state of oxidative stress is encountered, leading to excess generation of reactive oxygen species (ROS). The augmented production of ROS leads to oxidation of native LDL-C, resulting in the formation of ox-LDL.¹⁸ Oxidative stress also leads to expression of adhesion molecules which facilitate adhesion of monocytes and other inflammatory cells to the endothelial surface and their transmigration to the intima (Figure 1).^19,20 These monocytes subsequently mature into macrophages and then with absorption of LDL and ox-LDL transform into foam cells. The macrophages release large amounts of ROS and growth factors such as angiotensin II (Ang II), which reduce the activity of endothelial nitric oxide synthase (eNOS) resulting in low levels of constitutive NO. Ox-LDL also directly induces endothelial dysfunction by reducing eNOS activity and denaturing preformed NO. Ox-LDL directly and indirectly, via release of Ang II, also induces smooth muscle cell proliferation and migration to the intima subsequently resulting in altered shear stress and intima-media thickness (Figure 1).

Figure 1.

Interplay and crosstalk between elevated oxidative stress, an immature immune response, and altered renin–angiotensin–aldosterone system responses that can prime a preterm infant for development of cardiovascular disease later in life. Oxidative stress results in formation of ox-LDL which is taken up by macrophages which develop into foam cells. These cells release large amounts of reactive oxygen species and growth factors such as angiotensin II; the latter stimulates Type 1 receptor AT1R on endothelial cells and smooth muscle cells. Reactive oxygen species and AT1R activation repress endothelial nitric oxides synthase activity, resulting in reduced constitutive NO production and vasoconstriction. The immature immune response results in an imbalance of pro-inflammatory and anti-inflammatory cytokines, with enhancing effects on oxidative stress and vasoconstriction. Inflammatory cells under the influence of pro-inflammatory cytokines activate iNOS resulting in formation of large amounts of NO which can have a tissue injurious effect. Direct and possible long-term effects are listed at the bottom of the figure.

In preterm infants, the balance between production and elimination of ROS is deranged with enhanced production of ROS during the pre- and post-natal period.^21,22 Endothelial dysfunction which manifests as decreased flow-mediated vasodilation, higher pulse wave velocity, and increased intima-media thickness is often noted in the preterm population.^23,24 All these events are associated with a significant risk for development of CVD events.

Tauzin et al.²⁴ reported diminished whole-body arterial compliance and distensibility in an investigation of 5-day-old preterm infants that persisted at 7 weeks of life. In fact, these changes in vascular compliance may persist in young adults and contribute towards increased systolic blood pressure, thus posing a risk for CVD in adult life.²⁵

The Atherosclerosis Risk in Young Adults study investigated 524 young adults between 26 years and 30 years of age for aortic stiffness by measuring carotid-femoral pulse wave velocity. In this study, prematurity independently showed an inverse association of gestational age with pulse wave velocity.²⁵ Norman and Martin²⁶ complemented this data by showing a relationship between prematurity and endothelial dysfunction in 3-month-old infants. The loss of vascular compliance early in childhood may persist in young adults and contribute towards development of increased systolic blood pressure and as a potential risk factor for CVD in adult life.²⁷

Inflammation

Preterm neonates often present with an immature immune system, manifested as impaired production of certain cytokines.^28,29 Impaired cytokine release enhances oxidative stress, which induces a heightened state of inflammation with excess generation of pro-inflammatory cytokines and a reduction in anti-inflammatory cytokines (Figure 1). Intrauterine infection in the amniotic fluid (a common cause of preterm birth) can elicit a foetal inflammatory response characterized by elevated levels of C-reactive protein and pro-inflammatory interleukins (IL-1, IL-6, and IL-8), resulting in exposure of immature organs to a pro-inflammatory milieu.^29–31 Belderbos et al.³¹ showed a skewed pattern of toll-like receptor 4 (TRL4)-mediated cytokine production in neonatal blood, with a reduction in anti-inflammatory cytokines IL-10 and IL-12p70 in the 1st month of life.

The disturbed response to pathogenic stimuli may be the result of epigenetic modifications of haematopoietic cells during development.^32,33 NF-κB is a major transcription factor that controls transcription of many genes; changes in its activity have been linked to chronic inflammation and atherosclerosis.³⁴ The heightened state of inflammation ultimately leads to increased oxidative stress and endothelial dysfunction in preterm infants (Figure 1). Further, exposure to infections during neonatal intensive care stay, along with mechanical ventilation and post-natal excessive oxygen, puts the susceptible preterm newborn at risk of exaggerated inflammatory response. The abnormal inflammatory response, along with endothelial dysfunction and vascular hyper-reactivity may form the basis for considerable risk of CVD later in life (Figure 1).

Dyslipidaemia

The link between prematurity and a pro-atherogenic lipid profile has been studied by several investigators, but the results remain inconclusive.^14,15 A Finnish birth cohort study showed that 1-week longer gestation decreased total cholesterol from 6.7% to 6.2% on average, and apolipoprotein B (Apo-B) levels from 12.3% to ∼11%.¹⁴ Finken et al.¹⁶ observed a less favourable lipid profile to be strongly associated with high body mass index and with high LDL-C and Apo-B levels. As discussed in the previous section, formation of ox-LDL from LDL-C in a poorly regulated background is an initial and critical step in the pathogenesis of atherosclerosis. It may be hypothesized that an abnormal lipid profile and underlying augmented oxidative stress in preterm newborns accelerates or activates processes leading to the development of CVD later in life. The observed association of dyslipidaemia starting early in life and increased CVD risk during adulthood clearly has significant public health implications.³⁵ This becomes particularly important since there seems to be a significant relationship between LDL-C levels and CVD events.³⁵ There is a need to further explore the relationship of prematurity, gestational age, post-natal growth, and dyslipidaemia to CVD risk.

Post-natal growth and metabolic syndrome

Babies who are born preterm tend to have rapid post-natal weight gain. Many studies have documented that preterm birth is a major risk factor for development of metabolic syndrome in adult life.^6,10,14,15 Parkinson et al.³⁶ showed that preterm birth is associated with increased insulin resistance and high blood pressure, some of the major components of metabolic syndrome in adult life. Preterm infants who reach childhood often suffer from long-term adverse cardiac remodelling including dilation of ventricles resulting in globular cardiac shape, lower stroke volume with compensatory increase in heart rate, and increased carotid intima-media thickness.¹³ Cardiomyocyte development occurs early in gestation followed by endowment later, and prematurity is associated with linear increase in altered cardiac programming along with severity of growth restriction.^37,38 Inflammation and oxidative stress during the perinatal period of preterm babies leads to altered cardiomyocyte contractility and calcium signalling.³⁹ Despite these major and significant alterations associated with preterm birth, a meta-analysis did not find any major differences between outcomes of adults associated with metabolic syndrome that were either born full term or preterm.³⁶ Obviously, more studies are needed to address this association in the future.

Role of renin–angiotensin–aldosterone system

The renin–angiotensin–aldosterone system (RAAS) acts via multiple pathways to regulate blood pressure, intravascular fluid volume and the sodium-potassium balance. Of key importance is the hormone Ang II which acts on several receptors, such as Type 1 receptor (AT1R, Figure 1) which has received much attention. Of note, AT1R expression is up-regulated by ox-LDL. AT1R activation induces vasoconstriction and via aldosterone secretion enhances sodium and water retention. The RAAS is thought to promote atherosclerosis (directly or indirectly) by inducing vascular inflammation and remodelling, and furthering risk factors for hypertension, insulin resistance, and obesity.⁴⁰ Angiotensin II further promotes oxidative stress, endothelial dysfunction via NF-κB activation, expression of leucocyte adhesion molecules such as ICAM- and VCAM-1, and release of pro-inflammatory cytokines.^40–43 Angiotensin II also induces NADPH oxidase dependent ROS production which contributes to morbidities like hypertension and diabetes.⁴³ Angiotensin II activation leads to cardiac hypertrophy and regulation of cardiac capillary density via AT1R.⁴⁴ Angiotensin II binding to AT1R has been recently shown in vitro to induce autophagy in cardiomyocytes, an important mechanism of cell survival.⁴⁵ Prematurity has been associated with an increase in plasma renin and Ang II levels, as well as angiotensin converting enzyme activity.⁴⁴ After adjustment of confounding variables, a comparison between circulating and renal RAAS components in preterm infants born to pre-eclamptic and normotensive mothers identified differences that were attenuated by the body mass index of the individuals later in life.⁴⁶

Genetics and epigenetics

In rare cases, a genetic cause can be responsible for cardiovascular (CV) malfunctioning in preterm infants⁴⁷ but in the majority of cases no genetic cause can be identified. In the past decade, epigenetics has been considered as a potential mechanism that influences developmental origins of adult diseases. Epigenetic modifications of haematopoietic cells during development resulting in a disturbed response to pathogenic stimuli have already been mentioned.^32,33

DNA methyltransferases, histone modifying complexes, microRNAs, and in particular long non-coding RNAs (lncRNA) may singly or in combination be involved in gene expression and epigenetic regulation.⁴⁸ Two hallmark lncRNAs, Braveheart⁴⁹ and Fendrr⁵⁰ (short for Fetal-lethal non-coding developmental regulatory RNA) were found to be expressed in mouse embryonic stem cells and shown to regulate the expression of core gene regulatory networks involved in defining the fate of CV cells.^49,50 Embryonic heart development is regulated by lncRNAs,⁵¹ and lncRNAs are involved in regulation of NF-κB transcription throughout life, with downstream effects on immune responses and oxidative stress.⁵² It has been suggested that such lncRNA genes could be therapeutically targeted.⁵³ However, as yet there are no data to support a causal relationship between lncRNA abnormalities in preterm babies and cardiac development. Chronic prenatal hypoxia exerts a direct effect on epigenetic modification and programming of the gene coding for cardiac PKCepsilon; this appears to be linked to foetal hypoxia and vulnerability of the adult heart to ischaemic injury.⁵⁴ Valenzuela-Alcaraz et al.⁵⁵ recorded the cardiac function of 100 normal and 100 assisted reproductive technology-assisted pregnancies and followed the infants for 6 months. They reported that foetuses resulting from such technology, when compared with natural pregnancy foetuses, had a more globular heart with decreased longitudinal function and impaired relaxation with dilated atria.⁵⁵

Microbiome and atherosclerosis

Microorganisms present in the gut play a crucial role in absorbing and digesting nutrients from food and are indispensable for maturation of a healthy immune system. The gut microbiome develops after birth through a haphazard process, stabilizing after a few weeks and maturing into an adult-like microbiome during infancy.⁵⁶ The gut microbiome of neonates is affected by mode of delivery; for instance, caesarean section results in a gut microbiome containing more skin-type microorganisms than vaginal delivery.⁵⁷ Since premature newborns are often formula-fed, hospitalized and exposed to antibiotics, all these factors will have profound effects on their microbiome early in life. Whether any of these changes in the gut of low birthweight or small-for- gestational-age infants have long-term effects on CVD risk is not yet known.

The composition of the adult gut microbiome has been linked with various diseases, such as non-alcoholic steatohepatitis, inflammatory bowel disease, obesity, and cancer.⁵⁸ Shifts in the gut microbiome related to obesity include higher numbers of Firmicutes and, to a lesser extent.

Proteobacteria at the expense of Bacteroides.⁵⁹ In adult patients with CVD an abundance of Enterobacteriaceae and Streptococcus species has been reported.⁶⁰ Heart failure and ischaemic or dilated cardiomyopathy in adults are correlated with a less diverse gut microbiota, with decreased fractions of Coriobacteriaceae (members of the Actinobacteria phylum), and Erysipelotrichaceae and Ruminococcaceae (Firmicutes).⁶¹ Heart failure is also suggested to result from increased inflammation from bacterial translocation due to a disrupted gut epithelial barrier.⁶² An integral gut-blood barrier is essential, as disturbance can lead to a ‘leaky gut’ resulting in low-grade inflammation and endotoxaemia; both conditions have been linked to risk factors for CVD such as hypertension, obesity, insulin resistance, and chronic kidney disease,⁶³ though a correlation to premature delivery has not been studied to our knowledge. The gut microbiome of hypertensive patients contains fewer producers of short-chain fatty acids, e.g. Roseburia and Faecalibacterium species (Firmicutes).⁶⁴ In particular, reduced production of butyric acid is of relevance, since it has a CV health-promoting effect.⁶⁵ Moreover, increased total cholesterol levels correlated with enriched Prevotella and decreased Clostridium members in the gut.⁶⁶ Shifts in the gut microbiome can further result in changes of NO production, mostly by oral flora⁶⁷ or decreased microbial H₂S production; the latter has a vasodilating effect.⁶⁸ All these shifts have been linked to CVD in adults. One specific causative effect of the gut microbiome has been identified. As shown in Figure 2, the gut microbiome metabolizes quaternary amines including phosphatidylcholine, choline, and L-carnitine, which are converted to trimethylamine (TMA). This is oxidized in the liver to trimethylamine-N-oxide (TMAO), a bioactive compound which in return promotes atherosclerosis⁶⁹ and has prothrombotic effects.⁷⁰ CVD in adults typically results in increased production of TMAO, leading to accelerated atherosclerosis.⁷¹ While the biochemical and molecular mechanisms involving a role of TMAO in atherosclerosis and several other pathways are yet to be explored, TMAO and trimethyl lysine are associated with and unfavourable outcome in patients with CAD or ACS.⁷² However, currently there is little information available whether these also increase risks related to prematurity. The known shifts are summarized in Figure 2, although for most of these observed correlations the cause and effect relationship is unknown.

Figure 2.

Abnormalities in the gut microbiome are associated with various conditions that increase the risk of cardiovascular disease, including a decrease in short-chain fatty acid-producing bacteria, increased inflammation, and reduced H2S production. Changes in oral flora can result in less NO production. For several of these associations cause and effect are still obscure, with the exception of increased trimethylamine-producing bacteria. Trimethylamine results from bacterial digestion of dietary quaternary amines (choline or carnitine). In preterm infant, increased levels of trimethylamine-producers result in higher trimethylamine-N-oxide production in the liver. This increases the risk of cardiovascular disease via conditions indicated on the left. Abnormalities in the gut microbiome are associated with various conditions that increase the risk of cardiovascular disease, including a decrease in short-chain fatty acid producing bacteria, increased inflammation, and reduced H2S production. Changes in oral flora can result in less NO production. For several of these associations cause and effect are still obscure, with the exception of increased trimethylamine producing bacteria. Trimethylamine results from bacterial digestion of dietary quaternary amines (choline or carnitine). In preterm infant, increased levels of trimethylamine producers result in higher trimethylamine-N-oxide production in the liver. This increases the risk of cardiovascular disease via conditions indicated on the left.

Trimethylamine-N-oxide was also demonstrated to stimulate the generation of white adipose tissue.⁷³ The exact nature of the microbiome fraction responsible for TMA production is unknown and probably species from multiple taxa are involved.⁷⁴ One promising possibility to treat gut dysbiosis is the intake of probiotics, such as the mucin-degrading bacterium Akkermansia muciniphila, which in mice can prevent inflammation.⁷⁵ Interestingly, medications such as metformin stimulate growth of A. muciniphila and other species, so that the beneficial effect of this drug might, at least partly, be the result of microbial shifts.⁷⁶ Other probiotic examples that can lower the risk of CVD are Lactobacillus plantarum which has a cholesterol-lowering effect⁷⁷ or L. casei which can attenuate diabetes, at least in mice.⁷⁸ Such ‘smart probiotics’ could provide the ‘next-generation’ of beneficial microbes, administered to treat specific conditions.^79,80 Such strategy may be explored for treatment and prevention of CVD in high-risk infants, for instance by modulation of the gut microbiota at weaning.

Conclusions

Foetal programming of risk factors in premature babies that may result in early development of CVD in the adult is now well established. The underlying relationship between prematurity and adult CVD is influenced by prenatal as well as post-natal health, while nutrition, lifestyle, and environmental exposures also play a major role. The European Guidelines on prevention of CVD prevention in clinical practice also state that ‘history of premature birth is possibly associated with an increased risk of CVD in offspring (RR 1.5–2.0), which may be partially explained by an increased incidence of hypertension and DM’.⁸¹ The role of epigenetic dysregulation in premature CVD remains to be elucidated. Possibly, the adverse effect of one risk factor enhances the adverse effect of other risk factors. Further research is needed to investigate if some of these negative effects can be overcome by manipulation of the infant’s microbiome. With emerging evidence supporting the lasting effect of maternal health on adult outcomes of neonates, the significance of improving maternal health becomes even more important. Future research is urgently needed to understand the factors that influence the foetal epigenome and microbiome, thereby allow the development of early interventions to decrease the burden of CVD in adulthood. Survival rates of preterm neonates have increased much in the past decades, but poor outcomes from CVD later in life is of much concern. The concepts discussed here are merely suggestive as to how premature babies may develop CVD prematurely. These concepts need to be rigorously tested in order to enable treatment and prevention strategies that are based on scientific knowledge.

Funding

This work was supported in part by NIH/NIGMS [1P20GM121293].

Conflict of interest: none declared.