Obesity-related hypertension and its remission following gastric bypass surgery – A review of the mechanisms and predictive factors

Abstract

It is well established that hypertension and obesity appear to be associated. The exact mechanism by which they are linked is unclear and remains a topic of a great deal of research. Current NICE guidelines recommend that patients with a BMI in excess of 35 kg/m² should be considered for bariatric surgery if they have a concomitant obesity-associated condition, of which hypertension is one. The commonest bariatric procedure in the UK is the Roux-en-Y gastric bypass, which has been shown to result in long-standing remission of hypertension in up to 93% of patients. This paper summarizes the existing literature on the main theories as to how obesity leads to hypertension as well as the literature concerning the effects of gastric bypass surgery on hypertension.

Key Words::

• obesity
• hypertension
• gastric bypass

Introduction

The adverse health consequences of obesity are many and varied – almost every system in the body can be affected (1,2). A systematic review in 2010 suggested hazard ratios of between 1.1 and 5.1 for the development of cardiometabolic morbidity [type 2 diabetes mellitus (DM), hypertension, ischaemic heart disease and cerebrovascular disease] in adulthood amongst overweight or obese children as compared with children of normal weight (3). Attempts to quantify the exact reduction in life expectancy imposed by obesity have varied; however, the generally accepted figure of 7 years would imply that obesity has as great an impact as lifelong smoking (3–5).

The relationship between hypertension and adiposity has been well established in numerous studies over the past few decades across a variety of ethnicities and weight ranges (6–9). In 1987, Garrison et al. (8) estimated that up to 78% of cases of hypertension in men and 64% of cases in women may be directly attributable to patients either being overweight or obese (8,10). The connection between the two has been further clarified by such studies as Jones et al. (11), who showed that there was a linear relationship between blood pressure (BP) and body mass index (BMI) – for every BMI increase of 1 kg/m² in normal weight individuals the diastolic BP rose by 0.89 mmHg, whereas in overweight individuals it increased by 1 mmHg. Subsequent studies have shown a similar correlation (7). Having said all this, it is worth remembering that despite the increased prevalence of hypertension in the obese population, not all obese patients are hypertensive and not all hypertensive patients have a raised BMI (6,12). Although even modest BMI increases and decreases can result in corresponding changes in BP, in both the normal weight and overweight/ obese populations, there is still considerable inter-individual variability (6,12–14).

Pathophysiology of obesity-related hypertension

Exactly how obesity leads to hypertension is not fully understood; however, several overlapping theories have been proposed. Central to these theories is the concept that the adipose tissue of overweight or obese individuals is “dysfunctional” (10,15). This means that the tissue differs from normal adipose tissue by showing increased levels of adipocyte hypertrophy and macrophage infiltration and an altered secretory function of the adipocyte-released hormones (adipokines) (10,15). It is primarily this altered adipokine secretion component that underpins the hypotheses explaining the aetiology of obesity-associated hypertension.

Sympathetic nervous system dysfunction

The main proposed mechanism stems from the proven link between elevated body fat levels and overstimulation of the sympathetic nervous system (SNS) (6,12,16–18). Activation of the SNS may induce hypertension by inducing peripheral vasoconstriction, impairing the urinary excretion of sodium and water and by stimulation of the renin–angiotensin–aldosterone system (RAAS) (10). It should be borne in mind, however, that SNS activity may occur regionally or systemically and much of the literature on this topic is based on the erroneous assumption that systemic markers of SNS activity, such as those of skeletal muscle, are equal in effect to the SNS activity specific to those organs controlling BP, most notably the kidneys (12,19,20). It has been suggested that obesity may not in fact result in systemic SNS hyperactivity but may instead result in a more selective SNS activation process – for example, skeletal muscle and renal SNS activity may be increased in obese patients, but cardiac SNS activity may be reduced due to baroreflex inhibition with the increased heart rate seen instead being a consequence of reduced parasympathetic activity (12,20–22). Those studies conducted since the development of more site-specific neurochemical and neurophysiological techniques have been pivotal in enhancing our understanding of the pathophysiology of obesity-related hypertension (19).

As a result of these site-specific techniques, there is now good evidence clarifying the link between body fat levels and SNS activity (19,20). Several studies have shown that obese humans demonstrate up to twice the levels of post-ganglionic muscle SNS activity compared with non-obese subjects (13,17,19,20,23–25). It has also been shown that ethnicity plays a role in the relationship between SNS activity, obesity and BP (10,26). An example of this finding would be Weyer et al. (26), who found that muscle SNS activity positively related to body fat percentage in Caucasians in whom obesity and hypertension are both widely prevalent but not Pima Indians in whom levels of obesity are high but levels of hypertension are low. In addition, Alvarez et al. (23) showed that SNS activity is more closely related to levels of intra-abdominal (visceral) than total fat mass or abdominal subcutaneous fat independently of total body fat (6,12,19,23,27). This in turn would explain why such patients have also been found to have a greater association with hypertension and cardiovascular disease (6,12,23,27). The broad mechanisms linking obesity and SNS activity have been described in several review articles and are briefly outlined below.

Hyperleptinaemia

Leptin secretion by adipocytes occurs proportionately to fat mass and its levels are therefore raised in obese subjects (12,28–30). It acts primarily on receptors in the hypothalamus resulting in decreased appetite and increased peripheral thermogenesis (6,12,16). Epidemiological evidence, studies of patients with congenital leptin deficiency and animal studies have also suggested a role for leptin in renal SNS activation and increased BP in the long-term if not acutely (12,16,31–33). It appears to exert its central effects on the SNS by acting on receptors, which form parts of other CNS systems such as the melanocortin receptors in the anterior pituitary gland (12,34). It is thought that obese subjects may be selectively resistant to leptin's effects on weight control but not to its influence on renal SNS activity; however, the exact mechanism for this has yet to be proven (16,35).

Hypoadiponectinaemia

High molecular weight adiponectin has been shown to have a significant cardio-protective effect in terms of lowering BP and reducing the incidence of atheromatous plaque formation (12,36). In contrast to most adipocyte-secreted hormones, however, its levels in obesity are reduced (36–38). Exactly how adiponectin reduces BP is not known but studies on rats have shown a dose-dependent reduction in renal SNS function when given either intravenously or intraventricularly (12,39). In addition, adiponectin has been shown to stimulate the actions of endothelial nitric oxide synthase thus inducing a reduction in vascular tone and smooth muscle proliferation (10,36,40,41). The elevated levels of free fatty acid and tumour necrosis factor-α seen in obesity are thought to impair nitric oxide synthase function and thus contribute to increased BP in a parallel fashion to the effects of hypoadiponectinaemia (10,42).

Hypoghrelinaemia

Ghrelin produced in the stomach and pancreas increases during fasting and appears to trigger the sensation of hunger. Rodent studies have suggested that it also counteracts the effects of leptin on melanocortin receptors thus inhibiting renal and systemic SNS activity (43,44). Ghrelin has also been shown to increase endothelial nitric oxide production thus resulting in reduced vascular tone (45). Low ghrelin levels would therefore result in less central sympatho-inhibition and less systemic vasodilatation with a consequential increase in systemic BP.

Figure 1. Mechanisms by which obesity induces hypertension.

Insulin resistance/hyperinsulinaemia

Although the evidence implicating insulin resistance in obesity-related hypertension is fairly weak, it has been postulated that the two may be indirectly linked by the effects of the former in terms of arterial intimal damage or chronic lipid metabolism dysfunction (6). The main arguments against hyperinsulinaemia playing a role in hypertension in the acute or subacute settings are that studies of BP in animals in whom insulin has been infused intravenously or directly into the brain do not show a concomitant sustained rise in BP (6,12,46,47). Equally patients on therapeutic intravenous insulin drips or with proven insulinomas do not show a tendency towards hypertension (48,49).

Baroreflex dysfunction

Baroreflex function has been shown to be impaired in obesity, particularly visceral obesity, and there is evidence that this phenomenon is another action attributable to leptin (21,27,50–54). However, since the baroreceptor reflex acts primarily to maintain acute BP stability, there remains some doubt as to the degree of influence this mechanism may exert on the BP of obese individuals in the long-term (6,12).

Hypothalamic–pituitary axis (HPA) dysregulation

It has been suggested that simultaneous activation of the SNS and HPA may play a role in the development of obesity-related hypertension (6,55). Evidence to support this hypothesis came in 2001 when Grassi et al. (56) demonstrated that prolonged glucocorticoid administration resulted in SNS inhibition in obese but not normal weight subjects concluding that the HPA may affect SNS function in several ways (6,56). Since these early findings, it has been suggested that a variety of other metabolites, for example reactive oxygen species, may act on HPA pathways to induce sympatho-excitation with subsequent increases in BP (57).

Renin–angiotensin–aldosterone system dysfunction

The second major hypothesis linking obesity to hypertension, related to the overstimulation of the SNS, is the concomitant increased activity in obesity of the RAAS, which brings about rises in BP via a variety of hormones primarily by directly augmenting renal sodium and water reabsorption and systemic vascular tone (10,58,59). One of the major RAAS components, angiotensin II, has also been shown indirectly to increase BP acutely through central excitation of the SNS and baroreceptor reflexes (6,60–62). Paradoxically though, the effect on SNS function appears to reverse on chronic exposure (63). Although the RAAS hormones are mainly secreted from organs other than adipose tissue, nearly all of them are elevated in obesity and reduce in concentration following bodyweight loss (BWL) (10,58,64–69). It is clear that with the elevated levels of the RAAS hormones in obesity, a direct effect on BP could occur; however, the role of the RAAS on the SNS remains uncertain (10).

Figure 2. Factors associated with the resolution of hypertension following Roux-en-Y gastric bypass (RYGBP).

Systemic inflammation and oxidative stress

The third major hypothesis linking hypertension to overweight and obesity centres on the well established association between increased adiposity and elevated levels of systemic inflammation and oxidative stress (10,70,71). It is thought that the common link between the two is that several of the pro-inflammatory cytokines and acute phase reactants have been shown to impair the vasodilatory function of vascular mediators such as nitric oxide (10,70,72,73). In addition, the direct pressure effect of the adipose tissue on the kidneys is thought to elevate BP by encouraging sodium and water retention (64,74). Both pathways would provide plausible explanations as to why visceral obesity is particularly associated with hypertension as patients with predominantly elevated levels of intra-abdominal fat have been shown to have increased levels of systemic inflammatory markers, oxidative stress and intra-abdominal pressure as compared with those with predominantly subcutaneous fat (71,75–77).

Roux-en-Y gastric bypass (RYGBP) and hypertension resolution

In comparison with the number of papers investigating its effects on BMI and DM there is a relative dearth of data concerning the influence that RYGBP may have on BP. The largest series of RYGBP on hypertensive patients was reported in 2003 by Sugerman et al. (78), who investigated 1025 patients of whom 521 were hypertensive (defined as systolic BP ≥ 150 mmHg, diastolic BP ≥ 90 mmHg and/or the use of antihypertensive medication). At 1–2 years post-RYGBP, hypertension had resolved in 69% of these patients [excess BWL (eBWL) 66 ± 18%, 91% follow-up rate] with this figure falling to 66% at 5–7 years (eBWL 59 ± 24%, 50% follow-up rate) and 51% at 10–12 years (eBWL 52 ± 25%, 37% follow-up rate) (78). The risk factors for non- resolution of hypertension were increased age, lower eBWL and African-American ethnicity (p < 0.001, p < 0.001 and p < 0.02, respectively) (78). The, perhaps predictable, finding that hypertension is more likely to resolve or improve if more weight is lost echoes those of smaller similar studies (79,80). Interestingly, it has been shown that in the majority of post-RYGBP patients not only does BP tend towards normal levels but also that the natural circadian rhythm of BP is generally restored – a phenomenon that is typically impaired in obesity (81).

Hinojosa et al. (82) reported a mean eBWL of 66% in a cohort of 95 patients and a 46% complete resolution rate at 12 months post-RYGBP with a further 19% of patients showing some improvement in their hypertension. Analogous to the findings of several studies looking at factors associated with DM resolution Hinojosa et al. (82) found that the duration since diagnosis of hypertension was an independent risk factor for resolution with the complete resolution group having a mean duration of 53 months vs 95 months in the non-resolution group (p = 0.001).

The only other published factor associated with successful resolution of hypertension following RYGBP is that of DM status. Carbonell et al. (83) reported on a cohort of 3193 patients of whom 655 (20%) also had DM. Although the paper does not state how many of the 3193 patients were hypertensive preoperatively, it does conclude that those subjects without concomitant DM were significantly more likely to experience resolution of their hypertension than those with both conditions (74.4% vs 63.5%, p < 0.0001) (83).

Given that there are a variety of mechanisms by which obesity is thought to induce hypertension, it is perhaps not unreasonable to think that there may be a variety of mechanisms by which post-bariatric surgery BWL may induce its remission. It used to be thought that BWL itself was the key factor driving the return to normotension; however, a time-course analysis published in 2009 showed that improvements in BP occur well before any appreciable BWL (84,85). Ahmed et al. (84) found that in their cohort of 100 patients the systolic and diastolic BPs reduced by 9 and 7 mmHg respectively at 1 week post-RYGBP. These measurements fell by a further 6 and 2 mmHg, respectively, with an overall 88% hypertension resolution rate over the remaining 12-month follow-up period, suggesting that BWL itself is not the key determinant of BP normalization following RYGBP (84). It is likely that the early improvements in BP, if not the long-term improvements, are the result of a hormonal mechanism (84). One possible explanation was proposed by Sledzinski et al. in 2010 (86) who reported a 40% increased in serum nitric oxide levels 6 months post-vertical banded gastroplasty, although clearly the differences in the nature of this procedure and the RYGBP make it difficult to draw confident parallels. Improvements in endothelial vasomotor function and aortic elasticity following RYGBP have been described in the literature; however, the significance of these findings with regard to long-term BP control is not clear (87,88).

Conclusion

Whilst the link between obesity and hypertension is well established, there continues to be some controversy regarding the exact nature of the association. That BWL, whether surgical or non-surgical, effects a change towards normotension appears to be equally well accepted but again the mechanisms by which it does so remain elusive. In comparison with BWL and DM remission, few attempts appear to have been made to identify factors predictive of hypertension remission. Given the disease burden, this would seem to be a worthy area for future research.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.