A randomized, crossover study of the acute effects of acarbose and gastric distension, alone and combined, on postprandial blood pressure in healthy older adults

Postprandial hypotension (PPH), usually defined as a fall in systolic blood pressure (BP) of ≥20 mmHg, within 2 h of a meal [1, 2], occurs frequently, e.g. in 24–38% of ‘healthy’ older adults [3] and 25–40% of patients with type 2 diabetes (T2DM) [2, 4], and is associated with major adverse sequelae, including syncope and falls [1, 2], as well as increased mortality [3]. Current management is suboptimal [4].

The pathophysiology of PPH is heterogeneous, which may account for the lack of effective management. Multiple factors appear to be involved, including autonomic dysfunction, the release of gastrointestinal hormones, meal composition, gastric distension, and small intestinal nutrient delivery. After a meal, there is substantial splanchnic blood pooling [2], and in healthy young individuals, baroreflex mechanisms protect against a post-meal decrease in BP through compensatory increases in heart rate (HR), stroke volume and cardiac output [5]. In ‘healthy’ older adults and in patients with PPH, these compensatory responses are inadequate to maintain BP [2]. We have shown in healthy older adults and people with T2DM that the magnitude of the postprandial fall in BP is greater when the rate of gastric emptying [6], or direct small intestinal nutrient delivery [7, 8], is relatively more rapid. For example, when glucose is infused into the duodenum at rates spanning the normal physiological range, a 3 kcal/min load induces a much greater decrease in BP and rises in HR and superior mesenteric artery (SMA) blood flow than 1 kcal/min [8]. Moreover, PPH is associated with more rapid gastric emptying [9]. In contrast to the effect of small intestinal nutrient delivery, gastric distension appears to be protective in PPH. Drinking water with a meal has been reported to attenuate the postprandial fall in BP in healthy older adults [10, 11], as well as in patients with autonomic failure [10‐12], probably primarily by inducing gastric distension. For example, Gentilcore et at (2008) reported that gastric distension with water at a volume as low as 300 mL diminished the fall in BP in response to intraduodenal glucose in the healthy elderly [13]. The pressor response to water may also be greater in patients with autonomic failure [11].

The alpha-glucosidase inhibitor, acarbose, is used widely in the management of T2DM. By delaying intestinal disaccharide absorption, it reduces postprandial glycemia [14]. Acarbose also stimulates the release of the incretin hormone, glucagon-like peptide-1 (GLP-1) [15], and slows gastric emptying [16], which may contribute to glucose lowering in T2DM. We reported previously that acarbose (100 mg) attenuated the fall in systolic BP induced by oral sucrose in healthy older adults, an effect associated temporally with slowing of gastric emptying and stimulation of GLP-1 [17]. However, when administered intraduodenally i.e. bypassing any effect of slowing gastric emptying, acarbose also attenuated the fall in BP induced by sucrose [18], which may reflect an effect of acarbose to reduce splanchnic blood flow [18]. Subsequent studies support the efficacy of acarbose in the management of PPH to attenuate [19, 20], but not abolish, the fall in BP, although these have, for the main part, employed small cohorts and associated with substantial methodological limitations. Somewhat surprisingly, only one study has hitherto evaluated the effect of acarbose (50 mg) on the splanchnic blood flow response to a meal [21], which appears central to the hypotensive response. Furthermore, this study was in a heterogeneous group of individuals with PPH and was neither randomised nor blinded in design, compromising meaningful interpretation [21].

While it is apparent that gastric distension (induced by water drinking) and acarbose theoretically have complementary effects to attenuate the postprandial fall in BP, this has hitherto not been evaluated, which is surprising given the simplicity and safety of both approaches. The aim of this study was to determine the acute effects of water drinking and acarbose, alone and combined, on the BP and SMA blood flow responses to oral sucrose. To establish ‘proof-of-principle’, we studied a cohort of healthy older adults, rather than patients with PPH.

The studies were reasonably well tolerated. Flatulence and/or diarrhea were reported by 3 of the 10 subjects on acarbose days; in each case the onset of symptoms was about 3–4 h after ingestion. Three subjects fainted during the study (2 subjects on the distension day and 1 on the acarbose day between t = 60–75 min), which, in each case, resolved promptly after lying supine for 15–30 min. In 1 of the 3 subjects, fainting was concordant with PPH on the distension day. No subject had autonomic neuropathy (mean score: 1.4 ± 0.2). In 1 subject, the ultrasound images of SMA blood flow were suboptimal on all study days due to the presence of bowel gas, precluding their use. In another subject, insertion of the intravenous cannula for blood sampling was unsuccessful on all 4 study days due to poor venous access. Accordingly, BP and HR data are available in 10, while SMA blood flow and blood glucose data are available in 9, subjects.

Our study evaluated the acute effects of acarbose and water drinking, alone and combined, on the BP and SMA blood flow responses to oral sucrose in healthy older adults. We demonstrated that in this group, ingestion of sucrose induced a substantial increase in SMA blood flow and fall in BP, as predicted [17, 18]. These changes were essentially abolished by acute administration of acarbose in a dose of 100 mg; which was well tolerated. In contrast, water drinking alone (300 mL, 15 min before the sucrose load) to induce gastric distension had no significant effect on either BP or splanchnic blood flow and did not modify the response to acarbose.

This is the first study to evaluate the interaction between acarbose and gastric distension on ‘postprandial’ BP based on their potential for synergetic, or additive, effects. The observed responses to acarbose are consistent with previous reports, including our own [17, 18, 21, 35], and support the concept that acarbose will be useful in the management of PPH. We have further demonstrated that acarbose causes a profound attenuation of the rise in SMA blood flow induced by oral sucrose [18, 21] which is likely to be central to its anti-hypotensive effect. Such an effect was suggested in a previous study that was methodologically flawed [21]. The reduction in the splanchnic blood flow response to oral sucrose induced by acarbose is likely to reflect mechanisms unrelated to slowing of gastric emptying per se [17], given that the increase in SMA blood flow induced by intraduodenal administration of sucrose in healthy older adults is also markedly attenuated by acarbose [18]. The stimulation of GLP-1 as a result of the presence of nutrient in the more distal intestine may be relevant. In particular, we have demonstrated in healthy older adults that intravenous GLP-1 reduces the SMA blood flow response to intraduodenal glucose [36] and that the ‘short-acting’ GLP-1 agonist, lixisenatide, prevents the fall in systolic BP and reduces the rise in SMA flow following a 75 g oral glucose load in healthy older subjects and T2DM patients [37]. However, in relation to a role for endogenous GLP-1, it would be expected that stimulation of GLP-1 would be associated with a concurrent equivalent increase in glucagon-like peptide-2 (GLP-2) [38, 39], which would be expected to stimulate SMA blood flow [40]. The reduction in plasma glucose induced by acarbose is predictably associated with a reduction in plasma insulin, which has vasodilatory properties [1].

We were surprised that water drinking had no significant effect on either BP or SMA blood flow. There are a number of potential explanations. We selected a relatively low volume (300 mL) drink to optimise tolerability, influenced by the outcome of our previous study demonstrating that this intragastric volume of water markedly attenuated the hypotensive response to intraduodenal glucose in healthy older adults [13]. With our study design, it would be anticipated that the majority of water would have been emptied from the stomach in the 15 min before the ingestion of the sucrose. Furthermore, the effects of gastric distension are known to be volume-dependent [41]. Hence, we cannot discount the possibility that a larger volume of water, given immediately before the sucrose may have been more effective. Moreover, the pressor effect of water drinking may also reflect changes in plasma osmolality [10, 11]. While there was a substantial fall in BP in response to sucrose, we studied healthy older adults, rather than patients with PPH, and none had evidence of autonomic neuropathy, although assessment of autonomic nerve function was indirect. The latter may also be of relevance to the negative outcome given that the pressor response to water appears to be exaggerated in patients with autonomic impairment [11]. It should also be appreciated that the number of subjects we studied was relatively small and there were non-significant trends for minor effects of water drinking soon after the sucrose drink. Accordingly, a type 2 error cannot be excluded. In addition, GLP-1 was not measured, which may have provided mechanistic insights into the role of acarbose in modulating postprandial BP. A strength of our study was the inclusion of splanchnic blood flow measurement.

Previous studies have suggested that both acarbose and gastric distension may be potential approaches to reduce the postprandial fall in BP and our study aimed to evaluate the interaction between these interventions. The outcomes provide additional evidence to support the use of acarbose to attenuate the magnitude of the postprandial fall in BP, but not in combination with water drinking. The latter concept should, however, not be dismissed pending the outcome of further studies with greater subject numbers and particularly in patients with PPH and where water is consumed immediately prior to a meal.