The University of Copenhagen announced on November 10, 2026 that people who drink coffee more often tended to have a lower pH in the small intestine. Francesca Bucci and colleagues reached the finding by re-analyzing observational data from 61 healthy Danish adults. The peer-reviewed paper was published online on September 15 in the journal Gut Microbiology. The team combined food diaries with a swallowable measurement capsule that records pressure and transit time in the gut. To understand the results, it is important to distinguish between gut contractions, movement of contents, the urge to defecate, and actual bowel movements. The effect of coffee on the gut has varied depending on which of these is measured.
Food diaries and capsule measurements point to small-intestinal pH
The researchers analyzed data from PRIMA, a nine-day observational study conducted in 2021. Based on eight days of food records, they compared 13 people who drank coffee on 0 to 2 days with 48 people who drank it on 4 or more days. No one drank coffee on exactly 3 days. Average intake over the eight days was 17.7 mL per day in the low-frequency group and 343.5 mL per day in the high-frequency group.
Of the 61 participants, 50 ate a standardized breakfast at their first visit and swallowed a wireless capsule called SmartPill. The capsule measures temperature, pH, and pressure in the digestive tract. For example, when the capsule moves from the stomach into the duodenum, the pH rises sharply. The team used such changes to estimate the capsule's position, and used cues such as the temperature drop on exit from the body to determine the end of measurement. The main PRIMA report, published in 2024, used the same observational data to examine how pH and transit time in the gut relate to individual differences in gut microbiome composition and metabolism.
The analysis showed that the group that drank coffee on more days had a significantly lower, more acidic pH in the small intestine (p=0.022). People with higher average daily coffee intake also tended to have a lower pH, with a Spearman rank correlation of ρ=−0.42 (p=0.002). Figure 5B of the paper shows the pH distribution for each group as a box plot, but the text gives no specific median values. A comparison limited to the duodenum, the entrance to the small intestine, found no significant difference between groups. The measurements also do not show how the pH difference seen across the whole small intestine affects digestion or health.
Colon contractions and transit time gave different results
When Bucci and colleagues examined the capsule records in detail, people with stronger detection signals for four urinary caffeine metabolites, including paraxanthine, tended to have higher indicators of colon contraction. However, there was no significant difference in colon contractions between the two groups divided by coffee-drinking frequency. Nor was there any significant correlation between coffee intake or urinary caffeine-related indicators and colonic transit time. These results appear in the paper's results section and Figure 5.
| Capsule measurement | Indicator compared | Result |
|---|---|---|
| Colon contractions | Coffee intake frequency from food records | No significant difference between the two groups |
| Colon contractions | Urinary paraxanthine | Rank correlation ρ=0.31 (p=0.026) |
| Colonic transit time | Coffee intake and urinary caffeine-related indicators | No significant correlation |
These results are based on data from the 50 people who swallowed the capsule. The correlation analyses were exploratory, and no correction for multiple comparisons was applied to limit the chance of false positives from testing many items.
Paraxanthine is a metabolite produced when the body breaks down caffeine. The detection signal for urinary paraxanthine correlated with coffee intake in the food records (ρ=0.55 among the 61 people). The correlation remained even after excluding the 7 people who did not drink coffee at all during the eight days (ρ=0.42). However, paraxanthine was detected even in these 7 people, so urinary values may also reflect caffeine from sources other than coffee, such as tea.
Gut contractions include movements that mix the contents in place and movements in which contractions propagate in sequence to push contents onward. The capsule measured changes in pressure generated around it as it passed through the digestive tract. Because it did not simultaneously track the order of contractions at multiple locations, pressure changes alone cannot show how the contents moved. The authors suggest that movements that mix the contents may have increased, but they could not distinguish whether the increased contractions were involved in mixing or in transport.
Individual differences in the response right after drinking coffee
Earlier small experiments have examined how the gut responds immediately after coffee is drunk. In 1998, S. S. Rao and colleagues inserted a catheter with six sensors into the colons of 12 healthy adults and recorded pressure at multiple sites. Participants drank regular coffee, decaffeinated coffee, and water in random order, with volume and temperature matched across drinks. Compared with water, regular coffee produced a larger time-integrated pressure waveform (area under the curve) and more contractions propagating along the gut. According to the published abstract, however, there were no significant differences in the motor activity indicators between decaf and water, or between decaf and regular coffee.
In 1990, S. R. Brown and colleagues measured pressure in the rectum and sigmoid colon of 14 healthy adults. In the 8 people who said they usually feel an urge to defecate after drinking coffee, indicators of gut motility rose within 4 minutes of drinking either regular or decaf coffee. The other 6 showed no similar pressure response, and no response was seen in any of the 14 when they drank hot water. In that experiment, participants were divided by self-reports of their usual urge to defecate, and what was measured right after drinking was pressure change in the gut rather than actual bowel movements.
Associations with gut bacterial types and total amount
Bucci and colleagues also analyzed bacterial 16S rRNA genes in stool. The group that drank coffee more often had a greater number of detected bacterial taxa (p=0.038). However, the Shannon index, which reflects not only the number of types but also how evenly taxa are represented, showed no significant difference between groups (p=0.15). When they used flow cytometry to measure bacterial cells per gram of stool, there was no significant difference between the two groups (p=0.95). Cell counts did, however, correlate positively with the four urinary caffeine metabolites, and the correlation remained statistically significant after correction for multiple comparisons (q<0.05).
For overall gut bacterial composition, they examined the association with daily coffee intake separately from the comparison by frequency group. Bray–Curtis distance, which evaluates differences between stool samples based on the relative abundance of bacteria, showed no significant association (p=0.35). A test using Jaccard distance, which is based on the presence or absence of bacteria, gave a similar result.
Looking at individual bacteria, the relative abundance of the genus Lawsonibacter was positively rank-correlated with coffee intake (ρ=0.39, uncorrected p=0.002). However, when the two frequency groups were compared, no individual bacterial group showed a significant difference in relative abundance after correction for multiple comparisons. Separately, a 2024 study by Paolo Manghi and colleagues, which examined 22,867 people in the United States and United Kingdom with detailed dietary records, found that Lawsonibacter asaccharolyticus, a species in the genus Lawsonibacter, was associated with coffee consumption. In culture experiments, both regular and decaf coffee promoted the growth of this bacterium.
Results for short-chain fatty acids, which gut bacteria produce when breaking down dietary fiber and other substances, also differed by analysis method. When stool short-chain fatty acid concentrations were analyzed together, an exploratory association with average daily coffee intake appeared (p=0.02). In the comparison of the two frequency groups, however, there were no significant differences either in the overall combination of short-chain fatty acids (p=0.2) or in the concentrations of individual compounds. The amount of short-chain fatty acids in stool depends not only on how much is produced in the gut but also on how much is absorbed into the body. The stool analysis and small-intestinal pH measurements cannot identify which bacteria were involved in the small-intestinal pH difference.
Could coffee help relieve constipation?
In this study, the average age of the high-frequency coffee group was 40.9 years, compared with 32.8 years in the low-frequency group. The researchers also did not assign coffee intake to participants, and the type of coffee and the brewing method were not recorded. In the University of Copenhagen announcement, co-author Henrik Munch Roager named an intervention trial as the next research step: one that controls coffee intake and compares gut function between periods of drinking and not drinking coffee.
To examine whether coffee helps relieve constipation, participants would need to be assigned regular coffee, decaf, or a control beverage, with pressure changes in the gut right after drinking and subsequent bowel habits tracked separately. If, in addition to colonic transit time, the number of bowel movements over a set period, stool consistency, and degree of straining were defined in advance as endpoints, it would be possible to test whether coffee-induced gut contractions actually lead to symptom improvement.