Introduction

Cardiovascular diseases (CVD) are the leading cause of mortality, claiming an estimated 19.1 million lives annually1. CVDs comprise heart failure, ischaemic heart disease, coronary artery disease, cerebrovascular disease, and various illnesses of the heart and blood vessels2. More than four out of five CVD-related deaths are caused by heart attacks and strokes, and one-third of these untimely deaths occur in those under the age of 703,4. The most significant behavioural risk factors for cardiovascular disease and stroke include a poor diet, physical inactivity, cigarette use, and hazardous alcohol use2. The impacts of behavioural risk factors may manifest as elevated blood pressure, elevated blood glucose, elevated blood lipids, and excess weight and obesity in individuals. These “intermediate risk factors” can be tested in primary care facilities and suggest a higher risk of heart attack, stroke, heart failure, and other consequences. These modifiable risk factors act on non-modifiable risk factors such as age, ethnicity, or genetic factors, to increase the risk of CVD.

Those born preterm possess a non-modifiable risk factor for cardiovascular disease (CVD) that is incurred at birth. Preterm birth is the birth of an infant at less than 37 completed weeks of gestation5. Preterm birth per se, that is, shortened gestation in and of itself, has been linked with an elevated CVD risk profile for over 20 years6,7,8,9,10, and is considered an independent risk factor for CVD11,12. While the risk is greatest for those born at the limits of viability (~22 weeks’ gestation), even those born moderate to late preterm exhibit an elevated risk profile13. However, preterm birth remains a poorly recognised risk factor of CVD, even among those born preterm7,9,14.

Due to the combination of animal models and human clinical and epidemiological studies we now have a comprehensive understanding of the gross structural and functional alterations in the preterm cardiovascular system. When taken together, these offer a means for lifelong monitoring of this known non-modifiable cardiovascular disease risk factor that is preterm birth. The purpose of this review is: 1) to explore the known structural and functional factors contributing to cardiovascular dysfunction and CVD across the early lifespan, and 2) to propose that persistent structural alterations are the dominant cause of preterm cardiovascular dysfunction and disease risk throughout life.

Preterm trajectory of disease

Preterm birth is a complex and multifactorial phenomenon that can result from a combination of biological, environmental, and socioeconomic factors. In terms of biological factors, preterm birth can occur due to maternal medical conditions such as pre-eclampsia, infection, or cervical incompetence5. Additionally, genetic, and epigenetic factors can also play a role in preterm birth15,16,17. Furthermore, environmental factors such as socio-economic factors, poor nutrition, stress, exposure to pollutants, and substance abuse can also increase this risk of preterm birth5.

Preventive strategies for preterm birth include reducing risk factors such as maternal smoking and substance abuse, improving access to prenatal care, and promoting healthy behaviours such as good nutrition and stress management. However, preterm birth can have significant consequences for the infant, including an increased risk of neonatal morbidity and mortality, as well as long-term health consequences such as respiratory problems, cognitive and developmental delays, and an increased risk of chronic diseases in later life.

In recent years, prenatal impacts have been recognised as determinants of health and illness in later life, namely hypertension, ischaemic heart disease and heart failure18,19,20. Several epidemiological studies have demonstrated that prenatal and early childhood events may affect body composition and metabolism, thereby increasing the prevalence of several adult illnesses, including hypertension21,22, Type 2 diabetes23,24, and CVD13,25. Barker hypothesised that lower prenatal and postnatal growth may be associated with a higher risk of CVD in adulthood26. In addition, infants exposed to the Dutch Hunger Winter in early pregnancy during World War II were shown to have a higher incidence of obesity and cardiovascular disease later in life27. Following this, other studies established a definite association between preterm birth and CVD risk markers such as raised systolic and diastolic blood pressures21,28, impaired glucose tolerance and increased insulin resistance23,29,30,31, hypertriglyceridemia and low high-density lipoprotein levels in the blood32,33. Eriksson et al.34 looked at 4630 men born at Helsinki Hospital and found that men with a low ponderal index (the ratio of body weight to height) and slow weight gain in the first year of life had a greater risk of developing coronary heart disease later in life. Researchers found that premature babies were observed to experience ‘catch-up’ growth and had their body mass index go up between the ages of 1 and 12 were more likely to develop CVD34. However, this effect was only seen in children who had a ponderal index of 26 at birth34. Others have shown that low birth weight is a predictor of heart disease. However, low birthweight is an imprecise measure of growth in the womb and is not always caused by being born too early. Further research on preterm neonates has found that gestational age is a factor in the development of CVD13,19,35,36. In one study, men born preterm were shown to have greater quantities of total cholesterol, LDL-C, and apolipoprotein B than females33. Even after birthweight correction, these sexually dimorphic disparities in prematurely-born adolescents persisted33. When compared to full-term-born individuals, preterm birth was associated with higher LDL-C levels and elevated systolic and diastolic blood pressure33. Twin studies with dizygotic and monozygotic groups discovered that genetic and intrauterine environmental influences played a role in the development of CVD later in life17,37. Even though preterm birth is associated with a higher risk of developing CVD, the underlying processes or mechanisms that explain these correlations are not entirely known.

What is apparent is that this correlation with an elevated CVD risk profile is set from birth38,39,40,41. The capacity to adapt to the extrauterine environment determines survival in the immediate neonatal period, which has also been shown to have sexually dimorphic effects42,43. Male infants are born preterm at greater rates, exhibit more clinical complications during the neonatal period and are more likely to be readmitted following discharge than their female counterparts43,44. The cause of this discrepancy while unknown is likely multifactorial, with hormonal, genetic, and inflammatory factors playing key roles44. While more stable in the neonatal period, females born preterm exhibit an elevated risk profile throughout life with decompensation observable in adolescence33. While changes in neonatal care have significantly improved preterm survival, few improvements have eased the neonatal transition as much as the implementation of antenatal corticosteroids45. Indeed, while many infants now survive into adulthood without major comorbidities, all those born preterm carry a CVD risk inversely proportional to their gestational age13,20,46.

In addition to both maternal5 and neonatal factors28,34, antenatal corticosteroid (glucorticoid) treatments are increasingly associated with long-term disease outcomes47,48. Antenatal glucocorticoids have been routinely used since their introduction in the 1970’s to induce rapid lung maturation prior to birth45,47. While this treatment has become a mainstay treatment for prematurity – particularly at gestations <34 weeks (See Roberts et al.49) - the immediate effects on systemic growth may contribute to long-term cardiac, renal and insulin sensitivity47,50,51. Given the heterogeneity of the population, the long-term effects of antenatal corticosteroids remain conflicting, with late-preterm- and postnatal-administration appearing to add more controversy to this topic52,53,54.

Infancy

At birth, foetal proliferation and development of the heart and arteries abruptly slows, interrupting the normal process of cardiomyocyte differentiation in preparation for postnatal life55. Animal studies have demonstrated smaller hearts with reduced number of binucleated myocytes following preterm birth56 (Table 1). Foetal hyperplastic cardiomyocyte growth of cardiac tissue is ceased by the transition to neonatal life, potentially limiting the lifelong myocyte size and number38,48. This phenomena impacts both the left and right ventricle, contributing to altered geometry of the heart, as well as affecting the heart’s contractile function and overall performance20,57. Additionally, mechanistic studies of the preterm ovine heart have demonstrated diffuse collagen deposition seven times greater than in term hearts38,58, and studies in mice have shown that the presence of short, disorganised myofibrils that fail to align in the myocardium in preterm models59. These structural changes incurred because of preterm birth persist beyond infancy and ultimately determine a greater risk of CVD during later life20,60.

Table 1 Systematic review of preterm cardiovascular changes in comparison to term cohorts from infancy through to adulthood
Full size table

The functional complications associated with premature transition pose a more immediate clinical significance during the neonatal period61,62. Studies of piglets have demonstrated altered adrenoceptor profile in the neonatal period63, which when combined with excess sympathetic tone64,65,66 and altered circulating catecholamines67,68,69, contributes to impaired cardiac output and cardiovascular instability70,71. Instability which is further exacerbated by alterations in both pulmonary and systemic vasculature. Persistent pulmonary hypertension is three times more common among preterms, impairing right ventricle ejection fraction57,72. Growth arrest and increased stiffness of the aorta increases afterload on the heart, further impairing cardiac function39,73,74 (Table 1). Patent ductus arteriosus (PDA) in many preterm neonates (perhaps as many as 50%75) also impedes attempts to improve cardiovascular stability in the neonatal period76. Furthermore, microvascular networks of preterm infants are rarefied and disorganised76,77, and are typically maximally dilated at rest42,62,78,79,80,81. These complications pose significant problems for the clinician as inotropes can prove fickle in rectifying circulatory failure (40% fail to respond to dopamine or dobutamine71).

Reactivity tests including exposure to 4% CO2, hypoxia, and thermal or orthostatic stress have elicited responses that are contradictory, but importantly, consistently altered from that of term-born infants82,83,84,85 (Table 1). These altered responses to stress may be a symptom of the heterogeneity of the preterm condition at different gestational ages and under different levels of clinical severity. However, they may also provide critical insight into cardiovascular stability in the neonatal period. As demonstrated by Stark et al.62, microvascular perfusion in the immediate postnatal period correlates with both cardiovascular stability and mortality within 72 h of birth. Preterm infants with greater vascular flexibility, and thereby improved stability, tend to have reduced clinical severity and better outcomes. As such, while many structural alterations are present within the neonatal period (e.g., PDA, reduced heart and artery size), the functional responses to neonatal life, and therefore the functional complications, appear to be of greater significance to neonatal morbidity and mortality (Fig. 1).

Fig. 1: Relative contribution of structural and functional complications in preterm cardiovascular disease risk across the lifespan.
figure 1

Structural impairments (e.g., altered cardiac geometry38,146, ↓ microvascular density76,77) incurred with preterm birth ( < 37 weeks’ gestation) outside of clinically significant ones (e.g., patent ductus arteriosus) contribute relatively little to cardiovascular dysfunction, whereas functional complications (inotrope insensitivity71, ↑ microvascular perfusion62) are a significant cause of dysfunction (e.g., pulmonary hypertension96,147,148, ↓ cardiac output70,71). Surviving graduates of neonatal intensive care exhibit little cardiovascular dysfunction following discharge and throughout infancy. By childhood, structural limitations from prematurity become apparent (↓ LV mass94, ↑ RV mass95), resulting in some (dys)functional changes (↓ contractility94,95, ↑ BP90,93,145). Persistent structural limitations (altered cardiac geometry55,57,119, ↓ arterial diameter, ↓ microvascular density) contribute to cardiovascular remodelling (concentric hypertrophy57,112,134) and dysfunction in early adulthood (cardiac fibrosis58,134, arterial stiffness111,149, ↑ BP28). In combination with ‘traditional’ cardiovascular disease risk factors, these preterm-specific risk factors fuel the onset of disease.

Full size image

The acute transitional complications subside across the neonatal period as cardiovascular control matures65. Heart rate variability (HRV) studies have demonstrated maturation of cardiac control over the neonatal period such that by 2–3 years HRV is largely comparable between term and preterm infants64,65,66. Similarly, many of the functional issues such as aberrant dilation and cardiovascular instability resolve to a point where neonatal intensive care (e.g., inotropes, thermal support) is no longer required. Much of the clinical cardiovascular monitoring is frequently ceased following this neonatal period, with post-neonatal intensive care follow up focussing on critical neurodevelopmental milestones86,87.

Effects of preterm birth on cardiovascular health throughout life

By childhood, the shortened gestation becomes apparent in the structure of the cardiovascular system. Whereas cardiovascular control appears comparable between term and preterm infants64, increased circulating catecholamines88, alongside narrowed arteries73,89 and reduced microvascular density90, results in elevated blood pressure (BP)88,90 and altered stress responses in children born preterm91,92,93. These structural and functional cardiovascular alterations seldom reach clinical significance, particularly following moderate-to-late preterm birth, but may be early markers of future disease present in childhood36,94,95.

Both left and right heart geometry remain altered in childhood, impacting cardiac contractility36,94,95. Using echocardiography with extremely preterm-born children, Mohlkert et al.94 demonstrated significantly smaller left ventricles and impaired ventricular function, which is associated with a 4-fold higher risk of heart failure in children and adolescents born between 28 and 31 weeks’ gestation and 17-fold increase with gestations below 28 weeks36. Investigating the right heart, Mohlkert et al.95 also demonstrated increased right ventricle thickness and altered geometry alongside increased pulmonary vascular resistance. While unable to parse out differences between functional versus structural causation, the alterations in the preterm right ventricle are likely due to a combination of reduced or immature pulmonary vessels, and the resultant increase in pulmonary pressure96. A similar mechanism is likely at play in the systemic circulation. Certainly, a history of pulmonary neonatal diseases is associated with an elevated risk of pulmonary hypertension by adulthood72,97. In the systemic circulation, preterm-born children show reduced aortic, coronary and carotid artery diameters73,89, though this is significantly affected by the length of gestation. At later gestations carotid artery size appears comparable to term-born children98, indicating a threshold effect to arterial compromise in childhood. Combining the work of Szpinda99,100,101, Zhong et al.102, and Schubert et al.39, the aorta grows linearly in utero with elasticity increasing significantly from 31 weeks’ gestation (remaining similar between 20 and 31 weeks’102), but aortic growth abruptly slows at birth. Impaired growth of carotid and coronary arteries has also been observed in extremely preterm-born children73,89,98, suggesting that growth cessation is common among major vessels.

While elastin accumulation is maximal in the perinatal period103, its synthesis in the aorta is significantly impacted by intrauterine growth restriction104,105 and presumably also by prematurity74. Furthermore, as collagen and elastin content remains almost constant from infancy up to 3 years106,107, failure to synthesise adequate amounts of elastin due to premature birth may permanently impact arterial compliance103,106,108. Indeed, Odri Komazec et al.108 identified decreased elasticity and increased stiffness in aorta of preterm-born children (<32 weeks’ GA), and these characteristics have been similarly observed in adolescence and adulthood following gestations of 30-34 weeks’109,110,111 (discussed further below). The altered compliance in the major arteries of preterm children is only exacerbated by reduced arterial diameters and microvascular rarefaction90. This is likely the cause of elevated blood pressure observed by Bonamy et al.90. The elevations in BP, while minor in childhood ( ~ 4 mmHg), further drive cardiac maladaptation and the propensity for disease formation.

The persistence of altered cardiovascular structure presents clinically and epidemiologically in adulthood. Studies by Lewandowski and colleagues57,112 show that the morphometric changes observed in preterm infants and children persist into young adulthood, with magnetic resonance imaging revealing significant differences in both left and right ventricular structure. Functionally, in two meta-analyses of preterm-born young adults, preterm birth was associated with 4.2 mmHg113 and 3.4 mmHg114 elevations in systolic BP, respectively, with both analyses noting stronger effects in women. Additionally, a recent large-scale study by Crump13, identified an adjusted hazard ratio of 1.28 and 2.45 for new-onset hypertension in preterm- and extremely preterm-born adults (18-29 y/o), respectively. Similarly, Risnes et al.115 observed a 1.4-fold and 1.2-fold increase in mortality in early and late preterm born individuals between 15 and 50 years. Supporting this, Crump et al.28 observed a significant relationship between preterm birth and prescription of antihypertensive medications in young adults (25–37 y/o). Preterm birth has been further linked with heart failure36,116, ischaemic heart disease117,118, and pulmonary vascular disease97, though this risk is strongly—and inversely—related to gestation6,13. In a register-based cohort study, Carr et al.116 observed a 17-fold increased risk of heart failure in those born extremely preterm (<28 wks’ GA), with this reducing to 3.6-fold in those born very preterm (28-32 wks’ GA). In terms of ischaemic heart disease, Crump’s register-based cohort study observed a 53% increased relative risk of developing ischaemic heart disease in preterm born individuals aged 30-43 years117. Such findings in those born preterm are perhaps unsurprising given the continuity of cardiovascular dysfunction from infancy, adolescence, and adulthood.

A clear trajectory of decompensation can be observed through adolescence and adulthood, precipitated by persistent structural alterations in the heart and vasculature of preterm-born children. The structural limitations, such as altered cardiac geometry, narrowed and rarefied vasculature become more pronounced by adolescence (Table 1). The hearts and arteries of preterm-born adolescents are smaller (LV55,119; RV55; aorta119,120,121), resulting in greater vascular resistance121 and elevated BP33,109,121,122,123,124. Many studies, though not all98, have also observed greater arterial stiffness and intima-media thickness; however, the causal mechanisms remain unknown. While data is sparse in adolescents, the elevated BP and vascular resistance do not appear to impair cardiac output (LV function55,119) or vascular function10,122 and no signs of concentric hypertrophy can be observed at this age55. However, by adulthood, those born preterm exhibit hypertrophic and functionally impaired hearts, narrowed and stiffened arteries, vascular dysfunction and rarefaction (Table 1). While much of the evidence is observed at gestations below 29 weeks, structural and functional alterations consistent with the overarching pathology are observable at later gestations (~34 weeks). Indeed, the conventional risk factors for CVD in young adults born preterm are often present across the spectrum of prematurity13,108. This unique aetiology of preterm-related CVD has driven calls for clinical recognition14,18, and potentially a new cardiomyopathy36.

Stress reactivity

Stress tests are frequently employed to expose underlying cardiovascular dysfunction that is obscured at rest. Indeed, cardiopulmonary exercise testing is commonly used in the diagnosis of CVD. In populations with known risks of CVD, an impaired capacity to respond to – or recover from – the stressor may be indicative of early disease states. Stress testing may then provide useful prognostic insights into the preterm risk of CVD. Studies of those born preterm, from infancy through adulthood, have demonstrated persistent abnormal reactivity to a wide range of stressors, though there is ample room for expansion in these studies.

Autonomic and cardiovascular maturity has been examined in preterm infants using inotrope reactivity, hypercapnia, orthostasis and hypoxia (Table 1)82,83,84,85. Inotropes are routinely administered—with mixed efficacy—to improve cardiovascular compromise in hypotensive infants125,126. Mechanistic studies of preterm piglets have demonstrated reduced reactivity to both dopamine and dobutamine, with this possibly explained by immature cardiac and vascular adrenoceptor profiles (namely low abundance of cardiac ß1-adrenoceptors63 and vascular α-adrenoceptors71,84,127,128. As a result of immature neural control, preterm infants place greater reliance on circulating catecholamines; as demonstrated in the altered hypercapnic and orthostatic stress responses82,83, as well as hypoxic responses in preterm piglets84. Cohen et al. observed a 3- to 4-fold greater BP response compared to term-born counterparts with almost absent HR response when exposed to orthostatic stress83. Similarly, in response to hypoxia Eiby et al.84 observed a reduction in BP due to peripheral dilation with poor cardiac compensation. Together these studies demonstrate immature baro- and chemo-reflexes, particularly in the cardiac component of these reflexes. Notably, despite appearing stable at discharge from the hospital, these altered responses do not appear to resolve by term equivalent age82.

Childhood appears to be a deflection point in cardiovascular dysfunction when observed across the lifespan. Structural differences can be observed, but these appear to have a limited impact on function (Table 1). Exercise stress testing in children indicates a reduced exercise capacity, but due to a focus on respiratory function, limited inferences can be made to cardiovascular function91,92. One study in extremely preterm children exposed to acetylcholine challenge demonstrated elevated microvascular reactivity in children born appropriate for gestational age, though this only achieved significance compared to intrauterine growth-restricted preterm children and not term-born children93. Further studies in this age group would elucidate the impact of altered cardiovascular architecture and may prove beneficial for identifying early markers of disease.

By early adulthood, the structural limitations in the preterm cardiovascular system begin to produce pronounced dysfunction during stress testing. Using stress echocardiography, Huckstep et al.129 demonstrated progressive impairment in left ventricular ejection fraction and cardiac output during graded exercise, with this likely due to altered cardiac geometry exhibited at rest112,129. In a similar study, Macdonald et al.130 demonstrated impaired stroke volume augmentation and impaired right heart kinetics during exercise. This resulted in increased cardiac work for comparable stroke volumes and an increased reliance on heart rate response130. Importantly, these significant changes exhibited under stress were not present at rest130. Furthermore, examination of the vasculature during stress testing has shown increased stiffness in the form of increased pulse wave velocity, systolic blood pressure and pulse pressure in the brachial artery131. Increased vascular stiffness has been shown to increase afterload, and increase cardiac work for a given stress132, though the changes in preterm vasculature do not necessarily reach clinical significance133. This may explain the hypertrophic changes in both left and right ventricles57,112,134. Recovery, too, is impaired with preterm adolescents and adults exhibiting impaired heart rate recovery following graded exercise135,136. Heart rate recovery following exercise is primarily due to parasympathetic activation and sympathetic withdrawal137, and its impairment has been shown to be a predictor of cardiovascular disease135,137. Finally, in a cohort of preterm-born adults a 16 week exercise training intervention improved aerobic capacity and power, but not ambulatory systolic or diastolic blood pressure, a major risk factor for CVD138.

Collectively, the above studies demonstrate a trajectory of dysfunction as a result of persistently altered cardiovascular structure. The alterations in structure become progressively deleterious with age such that by early adulthood the altered cardiac structure exhibited at rest produces functional impairment under stress. Such studies demonstrate both the efficacy of stress testing in the preterm population as well as the value of stress testing as a prognostic test. While there are indications of system-wide dysfunction at rest, in the form of altered cardiac structure and vascular diameter and stiffness, the effects of these alterations appear dysfunctional under stress.

Current and future directions

As discussed in the outset, calls have been made for preterm birth to be recognised as a non-modifiable risk factor for cardiovascular disease for over 20 years now9,12,36,139. As a non-modifiable risk factor, short of preventing preterm birth, the root cause cannot be treated. Furthermore, the pathophysiological mechanisms contributing to CVD in preterm-born adults remains undetermined36,139. It appears, however, that structural insuffiencies strongly contribute to CVD risk (Fig. 1, Table 1), with contributing factors across multiple systems140. For graduates of neonatal intensive care, many are discharged from neonatal follow-up programmes early on, which consist mainly of neurodevelopmental milestones14. However, given the weight of evidence supporting lifelong risk of chronic disease including – but not limited to – CVD, and the absence of treatments, a sustained cardiometabolic follow-up programme offers a cost-effective and practical solution139,140. In ‘traditional’ CVD populations, acknowledging non-modifiable risk factors (e.g., family history), educating patients, and advising lifestyle interventions (e.g., diet, exercise, smoking cessation) are proven treatment options – used alone or in combination with pharmacological interventions to treat CVD141,142,143,144. Indeed, a recent questionnaire by Girard-Bock et al.14, found that many of the preterm-born adults were not even aware of their heightened CVD risk. They concluded that it is essential that long-term consequences of preterm birth are effectively communicated to preterm-born populations14. They, among others, noted that preventative strategies would be an effective treatment in the preterm population14,139,140. Current guidelines for BP management call for non-pharmacological management in patients with systolic BP of 120-139 mmHg140,144. While pharmacological treatment has been demonstrated to be effective in patients with systolic BP between 130-139 mmHg, it has not been recommended for young adults140,144. Elevated BP can be detected in the preterm population from childhood90,93,145, with dysfunctional traits manifesting in early adulthood (Fig. 1, Table 1). Jones et al.140, recommended heightened monitoring, including at-home BP measurement and early counselling on lifestyle interventions, with pharmacotherapy an option in high-risk patients. Given the efficacy of preventative strategies in other populations, such an approach will certainly save more in the long-run than waiting for the disease to manifest.

Conclusion

Events that alter the normal trajectory of early life development have profound implications for life-course health and wellbeing extending decades beyond the insult. Those born preterm are a heterogenous group in terms of sex, gestation, and neonatal morbidity. However, two things are clear: 1) preterm birth produces permanent structural changes to the heart and vasculature; and 2) preterm birth is associated with long-term risk of CVD.

Here, we have put forward the hypothesis, that the structural limitations incurred at birth produce adverse functional cardiovascular changes which, across the lifespan, drive maladaptive remodelling (e.g., concentric cardiac hypertrophy, arterial fibrosis) and CVD (Fig. 1). Such changes are pivotal stages in ‘traditional’ CVD aetiology. The key difference between the ‘traditional’ and preterm populations is that those born preterm require no further insult (e.g., poor diet, smoking, stress) to drive CVD, as the persistent structural changes drive hypertension, impaired cardiac output, and endothelial dysfunction.