Ayesha Shah ^1 2, Nicholas J Talley ³, Natasha Koloski ^1 2, Graeme A Macdonald ^1 2, Bradley J Kendall ^1 2, Erin R Shanahan ^1 2, Marjorie M Walker ³, Simon Keely ³, Michael P Jones ⁴, Mark Morrison ^1 5, Gerald J Holtmann ^1 2

Aliment Pharmacol Ther. 2020 Jul;52(1):155-167. doi: 10.1111/apt.15786. Epub 2020 May 15.

Author information

¹Department of Gastroenterology and Hepatology, Princess Alexandra Hospital, Brisbane, Qld, Australia.

²Faculty of Medicine and Faulty of Health and Behavioural Sciences, University of Queensland, Brisbane, Qld, Australia.

³Faculty of Health and Medicine, University of Newcastle, Callaghan, NSW, Australia.

⁴Psychology Department, Macquarie University, Ryde, NSW, Australia.

⁵Diamantina Institute, University of Queensland, Brisbane, Qld, Australia.

Abstract

Background: Small intestinal bacterial overgrowth may play a role in gastrointestinal and non-gastrointestinal diseases.

Aims: To use quantitative polymerase chain reaction (qPCR) to determine and compare bacterial loads of duodenal biopsies in asymptomatic controls, and patients with functional gastrointestinal disorders (FGIDs) and inflammatory bowel disease (IBD) including ulcerative colitis (UC) and Crohn's disease (CD). To define effects of gastric acid inhibition on bacterial load, explore links of bacterial load and gastrointestinal symptoms in response to a standardised nutrient challenge and compare bacterial load with glucose breath test results.

Methods: In 237 patients (63 controls, 84 FGID and 90 IBD), we collected mucosal samples under aseptic conditions during endoscopy extracted and total DNA. Bacterial load metric was calculated utilising qPCR measurements of the bacterial 16S rRNA gene, normalised to human beta-actin expression. Standard glucose breath test and nutrient challenge test were performed.

Results: The duodenal microbial load was higher in patients with FGID (0.22 ± 0.03) than controls (0.07 ± 0.05; P = 0.007) and patients with UC (0.01 ± 0.05) or CD (0.02 ± 0.09), (P = 0.0001). While patients treated with proton pump inhibitors (PPI) had significantly higher bacterial loads than non-users (P < 0.05), this did not explain differences between patient groups and controls. Bacterial load was significantly (r = 0.21, P < 0.016) associated with the symptom response to standardised nutrient challenge test. Methane, but not hydrogen values on glucose breath test were associated with bacterial load measured utilising qPCR.

Conclusions: Utilising qPCR, a diagnosis of FGID and treatment with PPI were independently associated with increased bacterial loads. Increased bacterial loads are associated with an augmented symptom response to a standardised nutrient challenge.