An artificial intelligence that analyses the microbiome of people with chronic constipation and recommends personalised diets may relieve symptoms more effectively than the standard treatment.
“The AI tells us what bacteria are and are not lacking in the microbiome of people with chronic constipation and recommends dietary changes to dietitians who then translate this data into personalised advice for the patient,” says at the Esrefpasa Municipality Hospital in Izmir, Turkey.
Advertisement
Chronic constipation is usually treated using laxatives that soften faeces orstimulate gut muscles to increase bowel movements. The condition can also be treated by injecting fluid into the bowel through the rectum, as well as via dietary changes, namely consuming more fluid and fibre.
Previous studies suggest that the microbiome of someone with chronic constipation differs from that of a person without the condition, suggesting that altering the microbiome could provide a longer-term solution, rather than temporary symptom-relief.
To test this idea, Tunali and his colleagues used data from previous studies to train an AI to analyse stool samples. The AI looked for bacteria that indicate a person’s overall microbiome and may be linked to any chronic constipation symptoms.
They then recruited 25 people with two or more symptoms of chronic constipation. These included straining on more than a quarter of defecations, passing hard or lumpy stools and having fewer than three spontaneous bowel movements a week.
For each participant, the AI analysed one stool sample. It then suggested dietary changes to a dietitian who devised a bespoke diet for each individual, which they followed for six weeks.
Another 20 people with chronic constipation acted as a control group, following a standard treatment for 10 weeks. This was made up of the laxative sodium picosulphate and a recommendation to consume more fibre and fluids.
Following treatment, the people who ate a personalised diet had an average 2.5-fold increase in weekly complete bowel movements. Those in the standard treatment group only saw their weekly complete bowel movements increase by a third.
The people who ate a personalised diet also reported a significant reduction in physical discomfort and were more satisfied with their bowel movements overall. They also saw a reduction in their “social discomfort”, defined as feeling anxious or embarrassed due to their condition.
In contrast, those who had the standard treatment reported being less stressed, upset and irritated by their condition, but had no further improvements.
However, the people who received the personalised diets had weekly appointments with a dietitian across the six weeks of their treatment, whereas those on the standard treatment only received a one-off recommendation to increase fluid and fibre uptake.
This suggests the benefits of the personalised diet may be due to regular dietitian appointments encouraging better adherence to a fibre- and fluid-rich eating plan, says at King’s College London.
Reference:
