Citric acid-enhanced Helicobacter pylori urease activity in vivo is unrelated to gastric emptying


Dr D. Y. Graham, Veterans Affairs Medical Center (111D), 2002 Holcombe Blvd., Houston, TX 77030, USA. E-mail:



In a previous study, the use of a citric acid test meal produced a rapid dose-dependent increase in urease activity that was significantly greater than that resulting from a pudding meal, ascorbic acid or sodium citrate. The mechanism was hypothesized to be related to the ability of citric acid to delay gastric emptying and possibly to enhance intragastric distribution of the urea.


To compare the effects of sodium citrate, two doses of citric acid and a pudding meal on gastric motor function.


Eleven normal healthy volunteers were investigated using non-invasive techniques to measure gastric emptying and gastric motility. We evaluated gastric emptying using the Meretek 13Ceebiscuit solid phase gastric emptying breath test, which employs a 340-calorie biscuit containing 200 mg of the edible 13C-blue–green alga Spirulina platensis, after the administration of test meals of pudding, 2 g and 4 g of citric acid and 2 g of sodium citrate. Electrogastrograms (Digitrapper EGG) were also recorded for 30 min before and 180 min after the test meal.


Gastric emptying, as assessed by the half-time (T1/2), was delayed similarly with the pudding (136.8 ± 9 min) and with 4 g of citric acid (144.5 ± 7 min) (> 0.7). Sodium citrate (108.7 ± 6 min) and 2 g of citric acid (110.1 ± 6 min) had similar effects on gastric emptying (P=0.986), and were significantly less effective in delaying gastric emptying (< 0.01) compared to pudding or 4 g of citric acid. The electrogastrograms remained normal and there were no differences among meals and no relation with the gastric emptying results.


The increased intragastric urea hydrolysis associated with citric acid test meals cannot be attributed to delayed gastric emptying. Changes in the intragastric distribution of urea or a direct effect of citric acid on the bacteria (e.g. via the cytoplasmic protein, UreI) are more likely to be responsible.


The original 13C-urea breath test employed a test meal in order to slow gastric emptying and to enhance the distribution of the substrate within the stomach to maximize the area and time of contact with the bacteria.1, 2 A number of nutrient and non-nutrient test meals have been suggested to accomplish these objectives.3–5 Recent studies have shown that citric acid test meals produce a greater degree of urea hydrolysis than traditional nutrient meals.6, 7 The effects of citric acid on the retardation of gastric emptying have been extensively studied and the results are consistent with citric acid retarding gastric emptying.8–12 For example, in a previous study, we showed a dose-dependent increase in gastric urease activity as the amount of citric acid was increased from 1 to 4 g (administered in 200 mL of water).13 In addition, the urease activity, as assessed by urea hydrolysis,14 was significantly higher with the citric acid test meal compared to the commercial pudding meal or a glucose polymer.13 We performed preliminary studies to investigate whether the citric acid-associated increase in urease activity was related to the low pH of the citric acid solution by comparing citric acid and ascorbic acid test meals. The ascorbic acid and citric acid test meals had similar pH values, but ascorbic acid was known to be about one-third as effective in slowing gastric emptying.8,12 In addition, we performed the urea breath test after the subcutaneous administration of pentagastrin as a second technique to address whether pH was a critical variable. Ascorbic acid and the subcutaneous administration of pentagastrin resulted in a significantly smaller increase in urease activity compared to 4 g of citric acid. We concluded that the Helicobacter pylori-enhanced urease activity with citric acid test meals was related to the marked effect of citric acid on gastric emptying and, possibly, the distribution of urea within the stomach.13 In this study, we tested these hypotheses by investigating the effect of citric acid on gastric emptying, both directly and on gastric electrical activity using the electrogastrogram.


We compared gastric emptying and electrogastrograms following ingestion of the US standard 13C-urea breath test pudding meal, 2 g and 4 g of citric acid and 2 g of sodium citrate. Tests were separated by at least 24 h.

Gastric emptying

We used the Meretek 13Ceebiscuit solid phase gastric emptying breath test to evaluate gastric emptying (Meretek Diagnostics, Nashville, TN, USA).15 This test employs a 340-calorie meal consisting of a 60-g rye roll, 30 g of cream cheese and 120 mL of white grape juice. The rye roll contains 200 mg of the edible blue–green alga Spirulina platensis which had been grown in an atmosphere of 99% 13CO2.

After an overnight fast, baseline breath samples were collected from the subjects. Before consuming the 13Ceebiscuit meal, the subjects received, in random order, one can of Ensure pudding (250 kcal), used in the US standard 13C-urea breath test, or 2 g or 4 g of citric acid, or 2 g of sodium citrate, in 50 mL of water. Breath samples were collected before and 30, 75, 90 and 180 min after the test meal.


In addition to the 13C-gastric emptying test, each individual underwent an electrogastrogram using three silver chloride electrogastrogram electrodes placed on the abdominal skin. The electrogastrogram signal was amplified using a portable electrogastrogram recorder (Digitrapper EGG, Synetics Medical Inc, Irving, TX, USA) with low and high cut-off frequencies of 1.8 and 15 cycles per minute (cpm), respectively. The electrogastrogram data were stored on the recorder and downloaded to a personal computer using the software Multigram Version 6.4 (Synetics Medical Inc, Irving, TX, USA). Analysis was based on the percentage of 2–4 cpm slow waves. The electrogastrogram was recorded for 30 min in the fasting state and for 180 min after the test meal.

Subject population

Normal volunteers were tested. Inclusion criteria included men or women between 18 and 75 years of age, inclusive, judged to be in acceptable health based upon the results of a medical history and physical examination. Exclusion criteria included participation in a drug study within 4 weeks before this study and clinical or suspected abnormalities in gastric emptying.

All subjects provided written voluntary informed consent. The protocol was approved by the Institutional Review Committees at Baylor College of Medicine and the Houston Veterans Affairs Medical Center.


The subject sample size was based on our previous trial,13 in which we showed clear differences in the effects of various test meals on the urea breath test. The differences among individuals were assessed using paired statistics (one-way analysis of variance for repeated measures). Pairwise multiple comparisons were performed using the Tukey test (SigmaStat v.2.03, SSPS, Chicago, IL, USA).

Breath samples were analysed for 13CO2 isotopic abundance using gas isotope ratio mass spectrometry. The increase in breath enrichment was combined with the estimate of CO2 production based on subject age, sex, height and weight to calculate mmol 13CO2/min. The lag in gastric emptying (Tlag) and the half-emptying time (T1/2) were calculated as described by Lee et al.15

The results of the electrogastrogram in relation to four types of electrical waves [> 1.8–2 cpm, 2–4 cpm (normal), 4–10 cpm and 10–15 cpm] were compared with the primary Tlag and T1/2 measurements. The electrogastrogram provided data on the gastric electrical activity and thus indirectly provided evidence about the gastric motility. The emptying test provided quantitative data regarding the rate of emptying of the gastric contents.

After deletion of motion artefacts, the electrogastrogram recording was divided into two portions: preprandial (in the fasting state before the test meal) and postprandial (after the meal). The power spectrum of each 1-min electrogastrogram was calculated and examined to determine the range of peak power. Bradygastria was defined as peak power within the range 1.8–2 cpm. Tachygastria was defined as peak power within the range 4–15 cpm. The percentage of normal 2–4 cpm was defined as the percentage time of 2–4 cpm slow waves over the entire observation period. This parameter reflects the regularity of gastric myoelectrical activity.


Gastric emptying, as assessed by T1/2, was similar with the pudding (136.8 ± 9 min) and 4 g of citric acid (144.5 ± 7 min) (> 0.7), and significantly longer (< 0.01) compared with 2 g of citric acid (110.1 ± 6 min) or 2 g of sodium citrate (108.7 ± 6 min) (Figure 1). Sodium citrate and 2 g of citric acid were similar (P=0.986).

Figure 1.

 Gastric emptying, whether assessed by the half-emptying time (T1/2) or the lag in gastric emptying (Tlag), was similar with the pudding and 4 g of citric acid test meals and, in both instances, was significantly greater than with 2 g of citric acid or 2 g of sodium citrate, which were in turn similar. *Significantly different from **. In addition, both T1/2 and Tlag bars with similar signs (e.g. *) were similar (i.e. > 0.7).

The results were similar when evaluating emptying using Tlag. Both pudding (64.7 ± 3 min) and 4 g of citric acid (73.6 ± 3 min) were similar (P=0.09), and greater (< 0.05) than either 2 g of citric acid (53.7 ± 4 min) or 2 g of sodium citrate (50.9 ± 3 min), which were similar (P=0.85) (Figure 1).

Electrogastrogram results were analysed as the percentage of normal 2–4 cpm gastric waves, 1.8–2 cpm bradygastria and 4–15 cpm tachygastria in time periods ranging from the first 10 min through the entire period. The electrogastrogram remained normal and there were no differences among meals and no relation with the gastric emptying results.


This study was designed to test directly the hypothesis that the increased gastric urease activity in vivo with citric acid test meals is related to gastric emptying. We used a direct test of gastric emptying to assess the ability of citric acid to retard gastric emptying, thereby increasing the duration of contact of 13C-urea with the bacteria in the stomach. We used the electrogastrogram technique in an attempt to investigate whether there was an alteration in gastric electrical activity which might imply a difference in distribution of 13C-urea within the stomach. We compared citric acid with sodium citrate which does not reduce the rate of gastric emptying.8, 12

We hypothesized that gastric emptying would differ between the pudding test meal and the citric acid test meal. We were unable to confirm this hypothesis, as the effects of citric acid and the pudding on gastric emptying were virtually identical. The rates of gastric emptying, whether measured by the lag in gastric emptying (Tlag) or the half-emptying time (T1/2), were significantly slower (< 0.01) for both the pudding and 4 g citric acid test meals than either the 2 g citric acid or 2 g sodium citrate meals. The expected dose response in gastric emptying with 2 or 4 g of citric acid was seen, confirming that the Meretek 13Ceebiscuit solid phase gastric emptying breath test was able to detect differences in gastric emptying. The electrogastrogram, which is an indirect measure of gastric motility, also remained normal during all the different test meals. No difference between test meals was seen, providing additional confirmation that enhanced urease activity with citric acid is unlikely to be entirely due to changes in gastric emptying or to the intragastric distribution of urea.13, 16–18

We conclude that the apparent increase in urease hydrolysis associated with the use of the citric acid test meal cannot be explained entirely by the ability of citric acid to delay gastric emptying. It has recently been recognized that a cytoplasmic protein, UreI, is present in H. pylori and is essential for infection and survival in acid.19 The activity of the intrabacterial urease increases as the pH is lowered, which is thought to be due to a change in the permeability of the inner membrane at acid pH, allowing urea access to the intrabacterial urease. The critical factor is currently thought to be UreI, which is a 21 K integral membrane protein, hypothesized to function as a proton gated urea channel, thus making urea accessible to the intrabacterial urease.20, 21 We hypothesize, based on the data presented here and in the previous study,13 that there is an interaction between citric acid and UreI in vivo that increases the access of urea to the urease enzyme. This will be the subject of subsequent experiments. pH per se appears unlikely to be the critical determinant, because pentagastrin stimulation of endogenous acid secretion had only a modest effect on urease activity that was not out of proportion to its effect on slowing gastric emptying. In addition, the pH was similar for each of the three doses of citric acid given in our previous experiment and for the ascorbic acid test meal, and yet the amount of urea hydrolysis differed markedly. Subsequent experiments will need to explain why there is a dose-response effect with citric acid. If the effect is through UreI, it is possible that ascorbic acid and citric acid have different effects on, or accessibility to, the postulated urea channel (i.e. UreI) that enhances urea permeability. The dose-response effect observed for citric acid also suggests that one variable might be the accessibility of the urease enhancers to the site where the majority of the bacteria reside. In vitro experiments will be needed to test these possibilities.


13Ceebiscuit solid phase gastric emptying breath tests were kindly provided by Meretek Diagnostics, Inc.