- Top of page
- Materials and methods
- Conflicts of interest
Abstract The aim of this study was to determine whether the nitric oxide (NO) synthase inhibitor, Ng-nitro-l-arginine-methyl-ester (l-NAME), reverses the effects of acute hyperglycaemia on gastric emptying and antropyloroduodenal (APD) motility. The study had a four-way randomized crossover (hyperglycaemia vs euglycaemia; l-NAME vs placebo) design in a clinical laboratory setting. Seven healthy volunteers [four males; age 30.3 ± 3.8 years; body mass index (BMI) 23.6 ± 1.2 kg m−2] were the study subjects. After positioning a transnasal manometry catheter across the pylorus, the blood glucose concentration was maintained at either 15 or 5 mmol L−1 using a glucose/insulin clamp. An intravenous infusion of l-NAME (180 μg kg−1 h−1) or placebo (0.9% saline) was commenced (T = −30 min) and continued for 150 min. At T = −2 min, subjects ingested a drink containing 50 g of glucose made up to 300 mL with water. Gastric emptying was measured using 3D ultrasound, and APD motility using manometry. Hyperglycaemia slowed gastric emptying (P < 0.05), and this effect was abolished by l-NAME. l-NAME had no effect on gastric emptying during euglycaemia. Hyperglycaemia suppressed fasting antral motility [motility index: 3.9 ± 0.8 (hyperglycaemia) vs 6.5 ± 0.6 (euglycaemia); P < 0.01]; l-NAME suppressed postprandial antral motility [motility index: 3.6 ± 0.2 (l-NAME) vs 5.1 ± 0.2 (placebo); P < 0.001]. Postprandial basal pyloric pressure was higher during hyperglycaemia (P < 0.001), and lower after administration of l-NAME (P < 0.001). Slowing of gastric emptying induced by hyperglycaemia is mediated by NO, and may involve the modulation of tonic pyloric activity.
Although delayed gastric emptying occurs in 30–50% of outpatients with longstanding type 1 and type 2 diabetes,1 its pathogenesis is poorly understood. Acute changes in the blood glucose concentration have major, reversible, effects on gastrointestinal motor function in both healthy individuals and patients with diabetes.2–5 Hyperglycaemia is associated with slowing of gastric emptying,2,3,5 reduced proximal gastric tone,4,6 inhibition of antral pressure waves,7,8 and stimulation of tonic and phasic pyloric contractions.9 However, the mechanisms mediating these effects are unknown.
Nitric oxide (NO) is a major inhibitory neurotransmitter in the gastrointestinal tract, and appears to act as the final common pathway mediating enteric smooth muscle relaxation.10 In healthy subjects, an increase in NO availability has been reported to slow gastric emptying.11–13 There is, however, no information as to whether nitrergic mechanisms mediate the effects of hyperglycaemia on gastric emptying and antropyloroduodenal motility.
We aimed to determine the effects of the NO synthase inhibitor, Ng-nitro-l-arginine-methyl-ester (l-NAME), on the delay in gastric emptying and changes in antropyloroduodenal motility, associated with hyperglycaemia in healthy humans. As glucose-induced insulin secretion is influenced by NO availability in vitro and in animal models,14–16 we also evaluated the effects of l-NAME on the secretion of insulin and the incretin hormone glucose-dependent insulinotropic polypeptide (GIP).
- Top of page
- Materials and methods
- Conflicts of interest
Our study has confirmed that acute hyperglycaemia delays gastric emptying of a nutrient liquid, and has established for the first time that this effect is mediated by NO. The dose of l-NAME used was based on previous studies20,28 and, as previously observed,20,28 changes in cardiovascular function were consistent with NO synthase inhibition. We cannot, however, exclude the possibility that some of the effects of l-NAME that we observed were not related to nitrergic mechanisms.29 Three-dimensional ultrasound was chosen to measure gastric emptying because of its superior accuracy over 2D ultrasound,21 and non-invasive nature, and because no ionizing radiation is required, unlike scintigraphy.22 Three-dimensional ultrasound has also been validated against scintigraphy for the measurement of gastric emptying of a liquid meal in healthy subjects.22 Although this method could potentially overestimate retention, as it is unable to differentiate between the liquid meal and gastric secretions, we do not believe that l-NAME had any major effect on gastric secretions, as the gastric volume was not significantly different between the two euglycaemic days. We acknowledge that we have studied only healthy subjects, and that our observations may not necessarily apply to patients with diabetes, in whom there may be changes to gastrointestinal signalling pathways secondary to either chronic hyperglycaemia or diabetes per se.
In this study, the marked slowing of gastric emptying observed with hyperglycaemia is consistent with previous reports.2,3,5,30 The observation that proximal, but not distal, gastric retention closely resembled total gastric emptying probably reflects the fact that the majority of the drink was retained, at least initially, in the proximal stomach. It should be recognized that our findings may not apply to the emptying of solid meals, although after ingestion of a mixed solid/liquid meal, solids also tend to remain in the proximal stomach until the majority of the liquid has emptied.31 The increased proximal gastric retention associated with hyperglycaemia in our study may indicate the suppression of fundic tone as a potential mechanism, although tone was not measured directly, which would have required the use of a barostat.32 Our observations regarding the effect of hyperglycaemia on fasting antral motility are also largely in keeping with the existing literature.7–9 Postprandially, however, we found that only short propagated antral wave sequences were suppressed during hyperglycaemia. This observation may reflect the fact that antral motility is normally suppressed after nutrient liquid ingestion,5,17,33 making any further difference between varying glycaemic states difficult to demonstrate. Although we did not find any effect of hyperglycaemia on the number of isolated pyloric pressure waves, in contrast to a previous report,9 we did observe an increase in basal pyloric pressure associated with hyperglycaemia postprandially, which has hitherto not been reported. Information on the effect of hyperglycaemia on proximal small intestinal motility is limited and inconsistent,19,34–36 but our study has not identified any major effect of hyperglycaemia on duodenal motility.
The role of NO in the gastropyloroduodenal region has been controversial. In humans, several studies involving either NO donors, such as intravenous nitroglycerin13 or sublingual glyceryl trinitrate,37 or inhibitors of NO production such as Ng-monomethyl-l-arginine (l-NMMA),11 indicate that NO mediates relaxation of the gastric fundus,37 slows gastric emptying,11–13,38–42 and decreases antral11,12 and pyloric13 contractions. Conversely, another study reported that l-NAME had no effect on antropyloroduodenal motility,28 while in the healthy elderly, l-NAME did not influence gastric emptying of a glucose drink.20 In animals, pyloric relaxation is impaired in neuronal nitric oxide synthase (nNOS)-deficient mice,43 and the relaxation of the rat gastric antrum appears to be dependent on nNOS expression,44 indicating an inhibitory effect of NO. However, other animal studies suggest that l-NAME is associated with slowing of gastric emptying.42,45 Furthermore, there may be sex-dependent differences in the contribution of nitrergic mechanisms to gastric motor function, with female rats having greater levels of nitrergic activity during health, and a greater propensity for disordered gastric motor function during diabetes, than male rats.44
In this study, the administration of l-NAME abolished the slowing of gastric emptying induced by hyperglycaemia, but had no effect on gastric emptying during euglycaemia, suggesting that the actions of NO may be glucose-dependent, which may partly account for the discrepancies in the literature regarding its actions, and indicates that in future studies the glycaemic state should be specified. The lack of effect of l-NAME on gastric emptying during euglycaemia, however, does not completely discount a potential role of NO in the normal feedback mechanism that regulates gastric emptying after a meal, particularly as the ‘physiological’ postprandial hyperglycaemia that normally occurs after a meal was prevented by the euglycaemic insulin/glucose clamp. We found no effect of l-NAME on antral motility during the fasting period, although the duration of observation was relatively short (i.e. 30 min). However, l-NAME suppressed all measures of postprandial antral motility. This contrasts with previous reports that increased NO availability was associated with a decrease, rather than an increase, in the postprandial antral motility index,11,12 although methodological differences may account for the discrepancy. For example, other studies either used a different NO synthase inhibitor (l-NMMA),11 or an NO donor,12 and one study used a semi-liquid, rather than a liquid, meal.11 The fact that l-NAME was associated with a suppression of postprandial antral motility, but not with a slowing of gastric emptying, suggests that antral motility probably does not play a major role in the emptying of a liquid meal. The number of isolated pyloric pressure waves was unaffected by l-NAME during both the fasting and postprandial periods. However, basal pyloric pressure was reduced by l-NAME postprandially, with a similar trend during fasting, indicating that l-NAME may have a suppressive effect on tonic, but not phasic, pyloric contractions. This could represent the dominant effect through which l-NAME abolishes the delay in gastric emptying induced by hyperglycaemia. Duodenal motility appeared little affected by l-NAME, suggesting limited involvement of nitrergic mechanisms in this region. Volunteers in our study were predominantly male, and in future studies relating to NO, it would be interesting to evaluate male and female responses separately.
The higher plasma insulin levels on the hyperglycaemic, compared with euglycaemic, days were anticipated. The lower incremental area under the insulin curve on the hyperglycaemia/l-NAME, compared with the hyperglycaemia/placebo, day, is modest, but is consistent with our previous observation that l-NAME is associated with less stimulation of insulin after oral glucose in healthy elderly subjects, compared with placebo,20 and implies that NO plays a role in mediating insulin secretion. The mechanism of this effect is unknown, but appears unlikely to involve the incretin hormone GIP, the levels of which did not differ between the study days, nor did our previous study indicate that glucagon-like peptide-1 (GLP-1) was involved.20 Given the latter observation, we did not measure GLP-1 in the current study. The involvement of NO in insulin secretion warrants further exploration, particularly in patients with diabetes, as both hyperglycaemia46 and diabetes47per se have been associated with impaired nitrergic activity. There are well-established links between NO and insulin release in animal and in vitro models, with one study suggesting that NO may contribute to glucose-stimulated insulin release,16 but others supporting an inhibitory effect of NO on beta-cell function.14,15 The lack of difference in plasma GIP between the four study days is somewhat surprising, given the markedly slower rate of gastric emptying on the hyperglycaemia/placebo day, compared with the other three days. We cannot provide a logical explanation for this. Hyperglycaemia and hyperinsulinaemia are unlikely to be involved in the secretion of GIP, given that levels on the hyperglycaemia/l-NAME day did not differ from the two euglycaemic days. Nitric oxide is also unlikely to be involved in GIP release, given the lack of difference in GIP levels between the euglycaemia/l-NAME and euglycaemia/placebo days.
In summary, we showed that the slowing of gastric emptying of a nutrient liquid, induced by acute hyperglycaemia, was abolished by the NO synthase inhibitor, l-NAME. Basal pyloric pressure was elevated during hyperglycaemia, and reduced by l-NAME. Insulin secretion appeared to be influenced by NOS inhibition. Further studies relating to the role of NO on gastric emptying and insulin release are now warranted in patients with diabetes.