Introduction

Hypertension remains one of the major risk factors for cardiovascular and cerebrovascular disease, and as such is one of the leading causes of morbidity and mortality worldwide [1]. There are considerable social and financial burdens placed upon people with hypertension, indeed healthcare costs in the USA alone are estimated at $131 billion annually [2]. Successful treatment of hypertension remains poor due to a mechanistic mismatch, with treatments targeting the established symptoms and not the upstream pathophysiologic root cause. Despite decades of research into its etiology, the vast majority of patients with hypertension maintain a diagnosis of “essential” hypertension, i.e., without known cause [3]. It has been argued that at least 50% of all cases of hypertension may have neurogenic origins, in particular relating to raised sympathetic nervous system activity [4].

In the last 30 years, there has been a growing body of epidemiological and experimental evidence showing that exposure to an adverse intrauterine environment during critical periods of development may predispose an individual to develop cardiovascular disease later in life. Indeed, it has been argued that many cases of primary hypertension may have a programmed or developmental origin [5, 6], with putative mechanisms including epigenetic modification, renal dysfunction associated with changes to nephron number, raised glucocorticoid levels, endothelial dysfunction, activation of the renin-angiotensin system (RAS), and autonomic dysfunction [7]. This review will present evidence regarding autonomic activity as a possible causative pathophysiological mechanism underlying the development of programmed hypertension.

Autonomic nervous system and hypertension

The autonomic nervous system has important functions in the control of blood pressure via direct regulation of cardiac output and total peripheral resistance, and indirect regulation of blood volume via alterations to renal function. In particular, elevated sympathetic nerve activity can increase blood pressure via changes to all of these variables, and is a commonly reported characteristic of hypertension [8,9,10]. Using both direct and indirect measures of sympathetic nerve activity, including microneurography and noradrenaline spillover there is convincing evidence that raised sympathetic nerve activity is present in both overt hypertension and in borderline hypertension [8, 9, 11,12,13]. Moreover, studies have found raised sympathetic nerve activity is present during childhood, prior to the development of clinical hypertension [10, 14], in normotensive subjects considered at increased risk of developing hypertension, such as children with a familial history of hypertension [15], those with white-coat hypertension [16] and during weight gain [17]. This suggests that, in many cases, raised sympathetic activity precedes clinical hypertension, supporting a causal role of sympathetic dysfunction in the etiology and pathophysiology of hypertension.

Studies also suggest that essential hypertension, and particularly resistant hypertension, is characterized by raised sympathetic cardiovascular drive [18,19,20], with both showing pronounced sympathoinhibition in response to anti-hypertensive interventions. Other studies indicate, elevated sympathetic nerve system activity precedes key predictors of cardiovascular disease outcomes, these include; obesity, carotid atherosclerosis, and left ventricular mass [21, 22]. Furthermore, on average, people with hypertensive disorders also have lower cardiac parasympathetic nerve activity [23] and higher heart rates [24]. Taken together, these data indicate that there is a strong neurogenic component associated with the pathophysiology and etiology of hypertension.

These clinical studies are supported by evidence from experimental studies, in particular those utilizing the spontaneous hypertensive rat (SHR), which have been important in establishing a causal link between sympathetic over-activity and hypertension. An early study by Judy and colleagues [25], showed that sympathetic nerve activity and blood pressure increased concomitantly and rapidly in juvenile SHRs as they aged, and by 5 weeks of age, both blood pressure and sympathetic nerve activity were significantly higher than values observed in normotensive strains of rats [25]. When sympathetic ganglionic transmission was reduced via hexamethonium administration in SHRs, both sympathetic nerve activity and mean arterial pressure were reduced to a level comparable with normotensive controls [25].

Autonomic nervous system and programmed hypertension

The mammalian autonomic nervous system begins its development during fetal life and continues after birth. Accordingly, both intrauterine and postnatal exposures may impact on the regulation of blood pressure. There is a wealth of epidemiological evidence indicating raised blood pressure can be influenced by in utero exposures to factors [7]. The earliest and by far most commonly reported observation is the inverse association between birth weight and risk of hypertension and cardiovascular disease in later life, such that those with the lowest birth weight are at highest risk of these cardio-metabolic diseases [26, 27]. In particular, those with fetal growth restriction, including mild to moderate forms, are at the highest risk of heart disease and type 2 diabetes, whilst those born preterm birth have evidence of raised blood pressure [28, 29]. These infants are exposed to distinct early life exposures that may confer risk via distinct pathways, including reduced time to develop prior to birth, and/or insufficient nutrient availability due to malnutrition or impaired utero-placental perfusion. To better understand the underlying mechanisms and the potential role for altered autonomic function, it is important to understand whether changes in autonomic function are due to prematurity or growth restriction alone and whether there is an additive effect in those born premature with low birth weight.

Autonomic function in individuals born growth restricted

A number of studies have investigated the relationship between fetal growth restriction and altered autonomic function, from the perinatal period, through childhood and to adult life. However, studies looking at growth restriction in the term born infants is limited. Using heart rate variability as an index for cardiac sympathetic activity, Galland and colleagues [30] found that in sleeping 1–3 month old term-born infants, those born small for gestational age (SGA) had increased cardiac sympathetic activity at rest and reduced cardiac reflexes in response to head up tilt, compared to normal birth weight infants.

There is evidence that raised sympathetic activity in those born SGA may persist throughout life. A prospective cohort study of 2089 five-year old children found that the cardiac pre-ejection period was shorter in those children born SGA, indicative of higher cardiac sympathetic activity [31]. Similarly, in adolescents, lower birth weight was associated with a shorter cardiac pre-ejection period at rest and in response to a variety of stressors [32]. This was true in both term and preterm births. Moreover, the reduced cardiac pre-ejection period accounted for 63–84% of the association between birth weight and raised blood pressure, supporting the hypothesis that raised sympathetic activity may play a role in the pathophysiology of hypertension linked to SGA. It is important to note that this study, which was performed in twins, found that within pair analysis showed an association between birth weight and pre-ejection period in dizygotic but not monozygotic twins, indicating that there are also significant genetic factors in the relationship between birth weight and raised sympathetic activity [32]. Finally, Boguszewski and colleagues [33], recorded sympathetic nerve activity directly using microneurography and found that young adults born SGA had a higher level of sympathetic nerve activity than those subjects born appropriate for gestational age. It has thus been proposed that augmented sympathetic activity may manifest during early development in those with impaired fetal growth, and that this, in turn, provides a plausible mechanism underlying the observed increased risk of hypertension in SGA individuals [33, 34].

Not all studies show evidence for increased sympathetic activity in people born small. Another microneurography study by Weitz et al. [35] found no differences in sympathetic activity in young adults born SGA compared to controls. A possible explanation for this is that, while both studies had similar levels of sympathetic activity in their SGA groups, the control group in the Weitz et al. (2003) study had higher sympathetic activity than the control group of the Boguszewski et al. [33] study. It is possible that the Weitz et al. (2003) control group also had raised sympathetic activity, due to high body fatness, that could account for their lack of a significant finding. Subjects in this study had a tendency towards overweight (mean BMI ≥ 26 kg/m2), with over a third of the control and SGA groups being clearly overweight (BMI > 28 kg/m2). In contrast, the subjects studied by Boguszewski et al. [33] were in the normal weight range (BMI 22–23 m2/kg). Body fatness is well known to be associated with increased resting sympathetic activity [36]. Finally, Weitz found no difference in blood pressure between low birth weight and healthy birth weight subjects [35], whereas, despite hypertension being an exclusion criterion in the Boguszewski study, there was a tendency for higher blood pressure in the SGA group. Therefore, although we cannot be certain as to why there are differences between the two studies, there is a clear question of selection criteria and possible confounders. Further carefully controlled studies can help to resolve this uncertainty.

Other studies have examined heart rate variability as a non-invasive index of autonomic function and found changes that cannot be attributed specifically to either the sympathetic or parasympathetic branch, but a decrease in overall cardiac autonomic regulation. A comprehensive examination of heart rate variability from 24 h ambulatory ECG monitoring found lower heart rate variability in 9 year old intrauterine growth restriction (IUGR) children ( < 2.5 kg) [37]. This was confirmed by Rakow and colleagues (2013), also in 9 year old children, who used frequency analysis to show decreased heart rate variability in total power and all frequency ranges in SGA [38]. Although these studies do not identify changes indicative of increased sympathetic activity, low heart rate variability is a predictor of incident hypertension and suggestive of generalized autonomic dysfunction [38, 39].

It is important to appreciate that birth weight percentiles used to define SGA is an imprecise measure of in utero adversity, representing a highly heterogenous group. It is therefore not surprising that any parameter, including sympathetic activity, will show a high degree of variability. Future studies utilizing better defined measures of in utero adversity will most likely achieve greater consistence within the study groups. Even so, results to date indicate that those born with impaired fetal growth have impaired autonomic control, with the majority of them showing increased sympathetic nerve activity. Programmed sympathetic hyperactivity may mechanistically link low birth weight with increased blood pressure, putatively predisposing these individuals to the development of hypertension and cardiovascular disease in later life.

Furthermore, in light of the obesity epidemic it is important to consider individuals at the other end of the birth weight spectrum, those born large for gestational age. A recent study by our group found reduced baroreflex sensitivity in infants with high body fatness compared to those with average body fatness. Furthermore, across the entire body fat spectrum there was a nonlinear association between newborn body fatness and baroreflex sensitivity which was independent of birth weight [40]. However, no differences were seen in blood pressure variability across the body fat spectrum [40].

Autonomic function in individuals born preterm

There is evidence that being born preterm affects autonomic regulation of cardiovascular function. Studies investigating cardiac autonomic control utilizing Wavelet transform of beat to beat interval found lower parasympathetic activity in week old premature infants compared to their full-term counterparts [41]. Follow-up studies examining heart rate variability show higher heart rates and reduced heart rate variability in preterm fetal growth restricted infants compared to preterm appropriate for gestational age controls on postnatal day one but not at 1 or 6 month post term age [42]. Adults born before 28 weeks of gestational age show reduced respiratory sinus arrhythmia (marker of responsive cardiac control) compared to their term counterparts [43].

Studies examining cardiac and vascular autonomic control utilizing heart rate variability and blood pressure variability show that preterm babies had reduced sympathetic vasomotor activity at 2–4 weeks age, and that at 5–6 months, both cardiac sympathetic and parasympathetic activity are lower during quiet sleep, but not during active sleep [44]. Similarly, other studies have shown, preterm infants (within the first week of life) exhibit reduced vascular autonomic control, quantified by a reduced baroreflex sensitivity [45] and have impaired normal maturational increase in baroreflex sensitivity [46]. The baroreflex is the most important regulatory mechanism of blood pressure homeostasis. Impaired sensitivity of the baroreflex in these individuals at an early age may place them at an increased risk of blood pressure lability and reduced capacity to buffer any pro-hypertensive provocations that arise. Studies in preterm, growth restricted infants show increased sympathetic response to a head-up tilt test [47, 48]. Although this seems inconsistent with the studies above that have reported decreases in sympathetic activity, it is possible that these infants have a normal, or even low sympathetic activity at rest, but an exaggerated sympathetic response to physiological (and possibly psychological) stressors.

This is consistent with a two hit hypothesis, whereby there is a programmed predisposition for hypertension that requires a “second hit”, in this case environmental stress, for hypertension to manifest. The predisposition appears in the form of susceptibility to stress.

Impaired fetal growth is an important risk factor for preterm birth; yet with potentially opposing effects on autonomic activity. Studies investigating prematurity per se found, day old infants born preterm with fetal growth restriction showed reduced cardiac autonomic activity as well as reduced cardiac sympathetic activity compared to their preterm appropriate for age counterparts. These differences seen at postnatal day one did not persist into infancy 1–6 months later [42]. A similar study in 5–12 year-old children by the same authors found preterm appropriate for gestational age children showed increased parasympathetic activation and blood pressure changes related to respiration compared to term appropriate for gestational age counterparts [49]. It appears these children have a heightened parasympathetic response to respiratory related blood pressure fluctuations, however this seems counterintuitive as increased parasympathetic activation is known to be protective of cardiovascular disease [50] with diminished parasympathetic control known to increase the risk of malignant cardiac arrhythmias and hypertension [43]. Other studies in children of similar age found an overall reduction in cardiac autonomic function [38, 51], the overall suppression of cardiac autonomic function may be indicative of an impaired capacity of this group to adapt to changes in the internal and external environment which may contribute to increase risk of later hypertension and cardiovascular disease.

Animal models of programmed hypertension

Following the strong associations made between low birth weight and increased cardiovascular risk, a number of animal models of programmed hypertension have been studied in the hope of elucidating the pathophysiological mechanisms.

Rat pups with low birth weight show significantly higher blood pressure as adults compared to their normal birth weight litter mates [52], illustrating a consistent neonatal phenotype with that of humans. A number of in utero insults have been found to reduce birth weight and program hypertension in animal models; maternal protein restriction, global food restriction, placental insufficiency and hypoxia, to name a few [53, 54].

Animal models examining the effects of prenatal undernutrition on offspring have found evidence for altered autonomic function with the resetting of the baroreflex to a higher set point, increased adrenaline levels, increase sympathetic and decreased parasympathetic nervous system activity [55]. Moreover, bilateral renal denervation completely abolishes hypertension in adult IUGR offspring [54], suggesting that increased renal sympathetic nerve activity contributes to the development and/or maintenance of hypertension in this model.

The exact mechanisms that link impaired fetal growth with changes in autonomic nervous system activity remain poorly described. It is unlikely that low birth weight is the cause of adverse subsequent health outcomes, but simply a surrogate observation that reflects the adverse in utero environment, as described above. In animal models, growth restriction is associated with increases in noradrenaline levels in the hypothalamus, and angiotensin and glucocorticoid receptor densities in the hypothalamus and brainstem, all of which are known to modulated sympathetic activity. There is also growing interest in the role of oxidative stress as many fetal insults are pro-oxidative [53]. Further elucidation of these putative mechanisms is crucial as potential early interventions, such as using angiotensin receptor blockers or antioxidants before the onset of hypertension may provide long term benefits [53].

Collectively a number of human and animal models demonstrate altered autonomic function following fetal growth restriction and or prematurity. In the context of human studies this is seen throughout the lifespan from infancy to adulthood, (Table 1). However, to date, it is unclear which branch of the autonomic nervous system is increase or decrease. These differences described may be attributed to different methodology employed to measure autonomic control over specific tissues as well as heterogeneity between study populations and participant characteristics. Furthermore, many studies detailing autonomic activity in small babies have been unable to specifically differentiate between the effects of growth restriction, constitutional smallness, and prematurity. Furthermore, birth weight is a widely used, albeit relatively crude, surrogate of fetal growth. Birth weight percentiles are unable to differentiate between the constitutionally small but well-nourished infant who has met their genetic growth potential, from an under-nourished infant of the same weight, whose intrauterine environment has restricted their growth trajectory. Therefore, defining alterations in fetal growth is problematic leading to heterogeneity within and between study populations. Future studies should take this into consideration. Furthermore, methodological differences between studies may also contribute to the discrepancies seen.

Table 1 Autonomic cardiovascular control in infants, children and adults born preterm or small for gestational age
Full size table

Indeed, preterm birth, constitutional smallness, and fetal growth restriction appear to have distinct associations with later hemodynamic and vascular health [28, 29, 56]. These differences may relate to the distinct etiologies, with implications for human populations in developing countries, where maternal malnutrition is an important contributor to fetal growth restriction, and in more developed nations where placental insufficiency is a relatively more important risk factor for impaired fetal growth.

Clinical implications

Hypertension is a complex condition, with multiple risk factors and pathophysiological pathways involved. The concept of developmentally programmed hypertension furthers our understanding of this disease yet adds to the complexity. Infants born growth restricted are among the tiniest and most vulnerable babies, and medical advances in the last decades have enabled the first generation of these babies to reach adulthood but leaving them with an increased risk of hypertension and later cardiovascular disease.

Countering a component of hypertensive disorders that has its origins in utero will open the door to the development of new strategies that prevent or specifically target these mechanisms. Nonetheless, the exact mechanisms by which in utero exposures alter autonomic activity and subsequent blood pressure remain relatively poorly described. The remarkable similarity of hemodynamic phenotype in offspring exposed to these distinct exposures suggests the possibility that there may be an overarching signal influencing programming in these studies.

It is apparent that neurogenic mechanisms underlie hypertension in a large number of patients and that optimal treatments are likely to be different for this group [57, 58]. However, currently there are no clinically available tests that differentiate neurogenic hypertension from non-neurogenic hypertension. The possibility of using non-invasive measures of autonomic activity, such as heart rate- and blood pressure variability, in a diagnostic setting should be investigated further.

Furthermore, low birth weight is a known risk factor for later hypertension with evidence for altered autonomic control. However, currently, there are no clinically established strategies to prevent cardiovascular events in those born growth restricted or preterm. Future studies should look to establish effective pharmacological and nutritional approaches in the management of blood pressure in those born growth restricted or preterm.