Delta-9-tetrahydrocannabinol delays the gastric emptying of solid food in humans: a double-blind, randomized study


McCallum The University of Kansas Medical Center, 3901 Rainbow Boulevard, Division of Gastroenterology & Hepatology, 4035-D, Kansas City, KS 66160–7350, USA.



Delta-9-tetrahydrocannabinol (THC), the active constituent of marijuana, is an effective agent in the prevention of chemotherapy-induced nausea and vomiting.


To determine the effect of THC on gastric emptying of a radiolabelled solid food in humans.


Thirteen healthy volunteers underwent gastric emptying studies after receiving THC and placebo in a randomized double-blind fashion on 2 separate days. THC, at a dose of 10 mg/m2 of body surface area, or placebo were administered.


Gastric emptying after THC was slower than placebo in all subjects. Mean percentage of isotope remaining in the stomach was significantly greater than after placebo from 30 min (85.5 ± 4.3% vs. 94.2 ± 1.4% placebo and THC, respectively, < 0.05) to 120 min (45.6 ± 7.2% vs. 73.9 ± 7.1% placebo and THC, respectively, < 0.001) after the test meal. No correlation was found between plasma THC levels and the delay in gastric emptying.


THC at a dose used for preventing chemotherapy-induced nausea and vomiting significantly delays gastric emptying of solid food in humans. Therefore, the anti-emetic property of THC may be mediated through the central nervous system.


Marijuana has been widely used for hundreds of years as an intoxicant or herbal remedy. Pure delta-9-tetrahydrocannabinol (THC) is the major active ingredient in marijuana (Cannabis sativa), and is one of 66 cannabinoid constituents of marijuana. It is an effective agent in the prevention of chemotherapy-induced nausea and vomiting.1 It has been shown to be superior to conventional anti-emetic agents such as prochlorperazine in numerous controlled studies.2[3]–4 Metoclopramide, a dopamine antagonist, is also effective as an anti-emetic agent in this setting.5 In addition to its strong central anti-emetic properties, metoclopramide also accelerates gastric emptying and gastrointestinal transit time.6 This property may play a complementary role in its anti-emetic action. We were intrigued by the possibility that THC may also affect gastric emptying in humans. To answer this question, we conducted a randomized, double-blind study to investigate the effect of THC on the gastric emptying of a radionuclide solid food in humans.


The study was approved by the Human Investigation Committee of the Yale University School of Medicine and all participants signed a consent form. Thirteen healthy volunteers, nine males and four females aged 19–30 years (mean age 22 years), participated in the study. All volunteers had experienced at least some exposure to marijuana in the past, an entrance criterion required by the Human Investigation Committee, but had not experienced any adverse reactions. All volunteers were in good health at the time of the study and had normal physical examinations. In particular, no subject was taking any medication that could have an effect on gastric emptying. The female volunteers were required to have a negative serum pregnancy test just prior to the study, and the gastric emptying evaluation was conducted in the first half of the menstrual cycle.

Each subject underwent two separate gastric emptying studies 3–7 days apart, using a radiolabelled solid test meal. They were instructed to abstain from marijuana use for at least 7 days prior to the study period. They also fasted and abstained from tobacco for at least 8 h prior to each gastric emptying study. One hour prior to the ingestion of the test meal, either the drug, THC, or a visually identical placebo was administered to each subject in a double-blind randomized manner. The dose of THC administered to the subjects was 10 mg/m2 of the body surface area, a dose often used in studies addressing the effectiveness of THC in prevention of chemotherapy-induced nausea and vomiting.3, 7

Gastric emptying studies

The gastric emptying study was performed as described previously.8 Briefly, after an 8-h fast subjects ingested a 275 kcal test meal consisting of a 7.5 oz can of commercial beef stew, two crackers and 100 mL of water. Technetium-99m sulphur colloid, 500 μCi, was injected into 30 g of cooked chicken liver which was then mixed with the beef stew. The meal was consumed over 15 min, and subjects were then placed supine under a gamma camera interfaced with a digital computer. The gastric region of interest was outlined by hand on the initial images by a nuclear medicine technologist. The counts were decay corrected and normalized to give percentage of solid meal retained in the stomach over time. Greater than 70% retention at 2 h represents 2 standard deviations above the mean for normal values established in our laboratory.

On each day a basal blood sample was drawn just prior to the administration of the drug or placebo. Hourly blood samples were then obtained, and hence four blood samples were drawn on each study day. The blood was centrifuged to separate the plasma, which was immediately transferred to a screw-cap glass tube and frozen. All specimens were sent in dry ice to the Center for Human Toxicology, Salt Lake City, Utah. They were analysed by gas chromatography/mass spectrophotometry9 for the level of THC and its metabolites by Dr Rodger Foltz through the courtesy of Dr Richard Hawks of the National Institute on Drug Abuse, Rockville, Maryland.

All 13 volunteers completed the study. The vital signs were monitored throughout the study days by one of the investigators. The subjects were interviewed both immediately after the gastric emptying study and 24 h later, for any adverse effects, specifically changes in mental state consistent with a ‘high’.

Continuous variables were evaluated for normality using the Wilk-Shapiro test and then appropriate parametric or non-parametric tests were applied. The paired t-test was used for parametric data and the Wilcoxon rank sum test for non-parametric data. Statistical significance was assigned for a P-value of < 0.05. The power of the performed test in the underlying population, with a sample size of n = 13, was 0.80 at α = 0.05. All data were expressed either as percentages or as mean ± S.E.M.10


Delta-9-tetrahydrocannabinol delayed the gastric emptying of solid food in each individual volunteer when compared to placebo. The mean gastric emptying data at different time intervals during the THC day and the placebo day are shown in Figure 1 . Compared to placebo, THC significantly slowed mean gastric emptying from 30 min (85.5 ± 4.3% vs. 94.2 ± 1.4% placebo and THC, respectively, < 0.05) to 2 h (45.6 ± 7.2% vs. 73.9 ± 7.1% placebo and THC, respectively, < 0.001) after the ingestion of the test meal. In eight out of the 13 volunteers the gastric emptying of solids was prolonged beyond the normal range (up to 70% at 2 h), thus achieving the criterion for an absolute delay in gastric emptying with potential symptomatic or clinical implications.

Figure 1.

. Comparison of gastric emptying of isotope-labelled solid food in human volunteers (n = 13) after the administration of 10 mg/m2 of body surface area of THC with placebo. THC is shown to significantly delay gastric emptying of solids from 30 min (< 0.05) to 2 h (< 0.001) after ingestion of the test meal.

The THC plasma levels achieved were in the same range as those recorded during the reports of THC as an effective anti-emetic in chemotherapy settings (Figure 2 2). There was no correlation between plasma THC levels and the degree of inhibition of gastric emptying. Mean plasma THC concentrations were observed to peak 2 h after administration of the drug, which coincided with 1 h after ingestion of test meal. When gastric emptying was still being significantly inhibited 2 h after meal ingestion, the mean THC levels remained at 80% of the peak concentration.

Figure 2.

. Plasma concentrations of THC after the administration of 10 mg/m2 of body surface area as determined by gas chromatography/mass spectrophotometry. THC concentrations are shown to peak 2 h after the administration of the drug, and remain elevated throughout the period of gastric emptying of the radionuclide test meal.

Twelve out of the 13 subjects experienced symptoms suggestive of a ‘high’. The one subject who did not experience the ‘high’ did have significant prolongation of gastric emptying. None of the volunteers noted any gastrointestinal symptoms or such specific adverse effects as nausea or vomiting.


The results of our study show that THC has a significant inhibitory effect on gastric emptying, and suggest the possibility that THC may have peripheral effects that do not play a role in its anti-emetic properties. It has been reported that THC inhibits gastrointestinal motility in mice.11, 12 Our present findings in humans would also suggest this and support the concept that one of THC’s peripheral effects is to inhibit gastrointestinal smooth muscle action. The present results are consistent with reports that THC inhibits smooth muscle function in other organ systems, causing relaxation of bronchial and vascular smooth muscle in addition to its effects on gastric smooth muscle.13

Bateman14 observed that THC has no effect on the gastric emptying of liquids in humans. However, in this study, the dose of THC used was smaller, and plasma levels were not measured. The gastric emptying of liquids was monitored by using real time ultrasound as opposed to the radionuclide method used in our study. Furthermore, because gastric emptying of liquids is mediated by a different mechanism from emptying of solids,15 their observation does not conflict with our results.

Delta-9-tetrahydrocannabinol-induced delay in the emptying of solids may be helpful in understanding some THC–drug interactions. For example, THC slows the absorption of alcohol and pentobarbitol.16 Although alcohol is absorbed to some extent from the stomach, it has to reach the small intestine to be fully absorbed, and a delay in gastric emptying slows the absorption of alcohol.17 Similar changes in the pharmacokinetics of other drugs, including orally administered anti-neoplastic agents, should also be considered. Because nausea is a psychological event that depends on sensory afferents such as pain,18 we can speculate that THC may reduce nausea by altering sensory input into the central nervous system. We intend to study the effect of placebo and THC upon gastric emptying with patients receiving chemotherapeutic agents in the future.


We wish to acknowledge the valuable assistance of Laurie Magyar RN, Gastrointestinal Assistant; Diane Errico and Robin Greene, Nuclear Medicine Research technologists; and Rita Severino and Kelly Crank for secretarial support in this study.