The sympathetic nervous system in hypertension

The importance of the sympathetic nervous system in short-term regulation of blood pressure is well accepted. Baroreflexes and central ‘command’ signals from higher brain centres dominate the moment-to-moment control of blood pressure through the sympathetic and parasympathetic nervous system. The question of whether neural systems participate in long-term regulation of blood pressure has long been debated [1], but remains an important question because there is increasing evidence for a role of sympathetic activity in a number of disease states such as hypertension [2]. Studies by Esler [3] have demonstrated that noradrenaline release from renal nerves is elevated in young borderline hypertensive patients and that they also have altered spillover of central monoamine neurotransmitters from subcortical regions of the brain. In many experimental models of hypertension, there appears to be an important contribution from the sympathetic nervous system, including the genetic forms such as the spontaneously hypertensive rat [4]. Neonatal sympathectomy of these rats produces a long-term blood pressure reduction [5], and interruption of brain neuromodulators or lesions of specific areas involved in sympathetic regulation prevents genetic hypertension as well as DOCA-salt hypertension [6–8]. Thus it is unsurprising that there has been a great deal of interest in the role of the sympathetic nervous system in hypertension, as there has been in other conditions such as heart failure [9] and, more recently, diabetes [10]. In humans, the techniques to measure sympathetic activity directly have been limited to skeletal muscle sympathetic activity. Apart from the difficulty of comparisons between patient groups, direct sympathetic recording of muscle only quantifies the frequency of the sympathetic nerve firing, which occurs with bursts of synchronous activity and relative silence in between bursts. The magnitude of the recorded burst is rarely measured because it reflects the recording conditions and is not an absolute or complete measure of sympathetic activity. Because the frequency is very strongly influenced by the heart rate and other factors, the interpretation of findings is often difficult. Furthermore, the differential regulation of regional sympathetic nerve activity means that measurements from one bed may not be reflected in other beds. Indirect measures of sympathetic activity have been developed, such as the noradrenaline spillover technique which involves the infusion of a small amount of radiolabelled noradrenaline [11]. The advantage of this method is that it is possible to sample from regional veins to measure sympathetic activity (indirectly) to a variety of organs independently, and this technique has been used extensively by Esler et al. [12] as well as many others [13]. It should be remembered that noradrenaline release appears to greatly exceed noradrenaline spillover from organs, and an assessment of regional sympathetic function may also require an understanding of the relationship of transmitter spillover to release [14]. Furthermore, the spillover technique is rather invasive, involving expensive radioactive materials and a highly skilled catheter insertion team, and is not therefore readily applicable to large numbers of patients. Assessment of sympathetic function using ganglionic blockade One of the more common ways to determine the contribution of the sympathetic nervous system to maintaining blood pressure has been to determine the reduction in blood pressure to ganglion blockade. The assumption is that quickly removing the sympathetic drive by blocking transmission through the ganglion will result in a blood pressure fall which is related to the degree of sympathetic activity. The technique itself is rather simple, requiring an intravenous catheter and measurement of blood pressure which can now be performed non-invasively. This has been used to suggest a greater sympathetic contribution to hypertension induced by high salt diet [15], blockade of nitric oxide [16,17] and cyclosporin [18], to name but a few. This approach has also been used extensively in humans by research groups interested in assessing sympathetic function. It is commonly used to investigate not only autonomic failure [19], but also to determine the changing contribution of the sympathetic nervous system under varying conditions such as ageing [20]. Although some investigators may consider this method as a ‘gold standard', it is far from perfect, and there has been much interest in finding a more reliable method. The difficulty is that differences in responses to ganglionic blockade between experimental conditions and between different patients do not always reflect a difference in sympathetic contribution. The change in blood pressure to ganglionic blockade is influenced by hormonal systems such as vasopressin and renin which can readily come into play when the sympathetic system is inhibited. Furthermore, changes to cardiac output can occur that complicate the interpretation of the change in blood pressure as being related to sympathetic vasomotor tone. Jones et al. [20] found that the differences between the response to ganglionic blockade with trimethaphan in healthy young and old men were due to differences in the cardiac output response to hypotension, rather than a difference in sympathetic vasomotor tone [20]. Thus, studies where the trimethaphan test is used exclusively could lead to incorrect conclusions. There are also patient groups that cannot safely tolerate large falls in blood pressure, such as those with heart failure. Sympathetic function assessed by blood pressure variability There has been a great deal of interest in the use of spectral analysis and other methods of determining variability in various frequency domains as an assessment of sympathetic activity. This method is particularly attractive due to its relative low cost, and its non-invasive nature when used with a Finapress device for measuring blood pressure. However, the enthusiasm with which this method has been embraced as an index of sympathetic activity [21] is equally matched by those who challenge its usefulness [22,23]. While blood pressure variability measured at low frequencies (< 0.15 Hz) is probably mediated by both sympathetic nerve activity and intrinsic vasomotor rhythmicity, the dynamic relationship between blood pressure and heart rate variability as quantified by transfer function analysis, reflecting the cardiac baroreflex itself, is mainly due to autonomic neural activity rather than other, non-neural factors [24]. Although, these indices are dependent on sympathetic activity, their magnitude reflects oscillations rather than the level of activity and, under some circumstances, they are not related. For example, in heart failure, there is increased cardiac sympathetic activity but a very much lower heart rate variability due to desensitization of β-adrenoceptors and lesser cardiac vagal tone [22]. Combinations of techniques Perhaps the safest way to assess sympathetic function is to use a variety of tests. Each of these may have their shortcomings but they are substantially strengthened when used together. This has been the approach of Diedrich et al. [25] as reported in this issue of the journal. The authors combined the decrease in blood pressure after ganglion blockade with low frequency oscillations in systolic blood pressure as a method of assessing the contribution of the sympathetic nervous system. They used four groups of patients: normals and hypertensives, as well as those with multiple system atrophy (MSA) and pure autonomic failure (PAF). The latter two groups have high and low sympathetic activity, respectively, and provided an excellent test concerning the usefulness of the estimates of sympathetic activity. The authors show that ganglionic blockade reduced blood pressure more in the hypertensive and MSA groups compared to normals and PAF groups. The reductions were accompanied by a reduction in low frequency power. There was a very strong correlation between the two methods of estimating sympathetic activity. However, the most interesting aspect of their study was the use of a dose–response curve to the ganglionic blockade to produce a regression between the fall in systolic blood pressure and low frequency power. This combined regression method produced slopes that could distinguish all four groups of patients. The MSA had the highest slope (approximately 30) compared to hypertensives (approximately 8), which was double that of normals (approximately 4). All these groups had strong correlation coefficients of greater than 0.86. By contrast, the PAF patients demonstrated no relationship between SBP and low frequency power. Thus, by calculating the ratio, a much more discriminating measure was evident. Clearly, this novel method would be very useful to determine the contribution of the sympathetic nervous system to hypertension and to phenotype patients with differing contributions from the sympathetic nervous system in maintaining blood pressure. One advantage is that it does not require complete ganglionic blockade or very large falls in blood pressure to produce the regression line. The method has not been tested to the point of defining normal and abnormal values, which could be used as discriminators, but, nevertheless, this work is a very useful step along this process, and hopefully will be taken up and expanded upon by others.