Blood pressure phenotypes in youth

The concept that evolution of biological organisms is directed by a complex interaction of genetic and environmental factors can also generally be applied to the evolution of human disease. In the last paragraph of the short but elaborate description of the fetal and infant origins of adult disease, the late David Barker challenges this notion by proposing that the ‘womb may be more important than the home’, suggesting that developmental factors in early life may be significant drivers that differentiate the pathway for health and disease in adults [1]. This publication was the initial formal declaration of the ‘Barker hypothesis’, resulting from epidemiological data collected largely for cardiovascular disease [2], and specifically, for development of hypertension later in life [3]. Although not immune to heated controversy [4], and notwithstanding that Barker's hypothesis was concerned essentially with in-utero development, this work has spawned a large number of studies addressing the importance of both fetal and early infant development for determinants of structure and function of biological systems in later life. The concept has been applied to a broad range of areas in human biology, from evolutionary associations of gap junctions, control of fundamental cellular functions of proliferation and differentiation and homeostatic regulation [5] to impaired early development of arterial elastin leading to increased arterial stiffness with age and to systemic hypertension [6]. The phenotypic evolution of arterial blood pressure (BP) progresses with age, such that in the elderly the majority of individuals have isolated systolic hypertension (ISH), that is, a wide pulse pressure (PP) [7]. In addition to humoral and neurogenic factors, the main cause of increase in PP in the elderly is largely related to increased stiffness of large conduit arteries [8]. The determinants of who develops hypertension in adult life are complex, with some as yet undetermined contribution of genetic factors, and a much larger contribution of environmental factors [9]. However, of all the risk factors contributing to cardiovascular disease, high BP is both one of the leading causes as well as the one with the greatest potential for modification. Obesity is one potential target for modification. The study by Lurbe et al.[10] in this issue of the Journal of Hypertension addresses the underlying factors contributing to the development of high BP by investigating a youth cohort to uncover BP phenotypes related to obesity. The study was conducted in 593 study participants (275 females) between 8 and 18 years of age (mean 12.3 ± 2.3 years). The study follows on from previous studies from the investigators addressing the effect of obesity and low birth weight (related to the Barker hypothesis) on phenotypes of BP in youth [11] and in the establishment of guidelines for management of elevated BP in children [12]. The investigation identifies BP phenotypes through measurement of office brachial cuff BP, central aortic pressure, PP amplification between the aorta and brachial artery, aortic pulse wave velocity, and 24-h ambulatory BP monitoring. According to established pediatric guidelines for BP [13], 511 (86.2%) were classed as normotensive, 48 (8.1%) with high–normal brachial systolic pressure, 24 (4.0%) with isolated systole hypertension (ISH) and 10 (1.7%) with systolic–diastolic hypertension (SDH). The relationship between central and brachial systolic pressure varied among the hypertensive phenotypes with central systolic pressure and PP amplification differentiating the ISH and SDH groups. Aortic pulse wave velocity showed the expected increase with BP. Ambulatory BP monitoring was able to confirm 90% of SDH cases and 75% of ISH. A significant feature of this present study [10], as well as of a small number of previous studies [11,14–16] is that central aortic BP was measured in children with devices that were developed for adults. Specifically, this study uses the SphygmoCor device (AtCor Medical, Sydney, Australia), which uses an invariant transfer function across all ages and BP applied to a radial or brachial pressure waveform to estimate the central aortic pressure [17]. The robustness of this technique is evident in the small errors in central systolic pressure when validated with invasive recordings of pulse waveforms [18] and in the fiducial tracking of central systolic pressure during large changes in systemic BP and heart rate as seen during Valsalva manoeuvres [19]. Although definitive large-scale outcome studies on the prognostic value of central aortic BP at the population level are eagerly awaited, the field has expanded to suggest a potential role of central aortic pressure for improved management of high BP as an underlying factor for cardiovascular risk [20]. As with conventional measurements of brachial cuff pressure, absolute values of central aortic pressure depend on the reliability of the noninvasive calibration [21]. However, by having the central and peripheral waveforms, calibration-independent metrics can be computed such as augmentation index and pulse amplification between central and peripheral sites [21–23]. The numerous studies now available have produced age-related reference values for central aortic systolic pressure and amplification in relation to cardiovascular risk factors [24]. Since the development of noninvasive techniques for estimation of central aortic pressure from the peripheral pulse over 20 years ago, the general recommendations have been that it be used in adults, that is, in study participants older than 18 years of age. The question then remains as to how should the findings of the investigation of Lurbe et al.[10] in young obese study participants be interpreted. Clearly, given the underlying causes of cardiovascular disease related to early developmental and environmental factors, there is a drive for improved cardiovascular assessment in children [25,26]. Importantly, the detection of significant prevalence of left ventricular hypertrophy in children with high BP [15,16,27] suggests a potentially significant role of central aortic pressure, as the difference between central systolic pressure is greatest in youth, and so central systolic pressure may be a better index of peak load on the ejecting ventricle. Hence, in the absence of validated systems (e.g. transfer functions) in children, taking into account the significant changes in body size with age (a factor less present in adults), a number of investigators have attempted to use the devices developed for adults in children. Close inspection of some of these studies shows that the low frequency part of the aortic-radial/brachial transfer function (0–4 Hz) is reasonably consistent, with a peak value of the modulus in the region of that for the adult transfer function (in the vicinity of 4 Hz) [15,17,28]. This suggests a potential explanation for the reported close agreement in estimations of central aortic pressure in studies performing invasive and noninvasive comparisons, or using the carotid artery waveform as a surrogate for aortic pressure [29–32], although some studies are in a small number of individuals. However, the majority of these studies report agreement for systolic pressure values, with minimal comparison of the reliability of the pulse waveform. In the context of interpreting the results of the study by Lurbe et al.[10] (and those of similar studies) anthropometric factors associated with the age of the cohort become a relevant consideration. The mean age for this study was 12.3 ± 2.3 years. Data for European children shows that for this mean age the mean height is 89% of adult height (age above 18 years) [33]. From echocardiographic data in a North American cohort, the ascending aortic diameter at age 12 years is 87% that of adults [34]. This suggests that arterial path lengths and vessel dimensions are comparable to those of adults, and factors governing pulse wave propagation would have minimal effects. A factor that may play a role is the distensibility of the arterial vasculature, resulting in lower pulse wave velocity in central conduit arteries. Indeed, this factor has been proposed as a potential explanation for detection of spurious hypertension in youth [35,36], although the concept has been challenged based on phenotypic differences of BP and stroke volume in youth [37]. Although the above explanation may be relevant for those above 12 years of age in the cohort, the errors would be clearly greater in younger study participants. We have quantified this in a group of 218 healthy, prepubescent children of 8 years of age (113 males) [30]. All had tonometric radial and carotid arterial pressure waveforms recorded using the SphygmoCor device. Carotid pressure was used as surrogate for central aortic pressure as reported previously for this cohort [29]. Central aortic pressure was estimated using the adult transfer function in SphygmoCor and compared with the carotid measurement, calibrated to mean and diastolic brachial cuff pressure. The estimate of central aortic systolic pressure using the adult transfer function showed a good correlation with measured carotid systolic pressure (r2 = 0.88) with a mean difference of 3.2 ± 2.6 mmHg. However, the correlation for waveform features defined by augmentation index was poor (r2 = 0.19) with a mean difference of 24.5 ± 10.4% [30]. The study by Lurbe et al.[10] provides useful information on the use of noninvasive means of assessing BP phenotypes in youth using measures of central aortic pressure. However, extrapolations of these measurements to meaningful physiological correlates are highly sensitive to age, particularly in the age range where there are large differences in the growth phase. Definitive studies will require the use of age and growth-dependent models relating central and peripheral BP. ACKNOWLEDGEMENTS Conflicts of interest There are no conflicts of interest.