Editorial: Association of Antibiotic Exposure With Microscopic Colitis—Authors' Reply

Abstract

We thank Drs. Tome and Pardi for their editorial on our study examining the association between antibiotic exposure and microscopic colitis (MC) [1]. We appreciate their careful literature review and insightful discussion regarding potential confounding factors that may influence the observed relationship between drug exposures and MC.

As the editorial highlighted, studies investigating medications associated with MC have consistently encountered challenges in disentangling causation from confounding. We aimed to address this through a self-controlled design [2], in which cases acted as their own controls, thereby mitigating important confounders, such as genetic predisposition, that are difficult to account for in traditional cohort studies [3]. In studies where external controls are necessary, matching by clinical characteristics and healthcare engagement patterns can further reduce confounding, thereby strengthening causal inference. However, we acknowledge that detection bias remains a consideration. Patients experiencing gastrointestinal symptoms on certain medications (e.g., antibiotics) may be more likely to undergo endoscopic evaluation, thereby increasing MC detection.

Medication exposures often occur in combination, particularly in older adults. These combinations (e.g., antibiotics combined with non-steroidal anti-inflammatory drugs or proton pump inhibitors) may amplify gastrointestinal side effects. The interplay of multiple drugs further complicates the challenge of establishing a direct causal link between any single medication and MC. Future research should investigate the impact of polypharmacy on MC incidence and severity, ideally through prospective, longitudinal studies with a rigorous selection of control groups.

Looking ahead, comprehensive studies are needed to understand how various drugs alter the intestinal environment, contributing to MC development. One potential option is investigating the gut microbiome's role and how antibiotic-induced dysbiosis may prime the colonic mucosa for an aberrant immune response. Incorporating stool metagenomics, mucosal immunologic markers, and detailed pharmacokinetic/pharmacodynamic data into prospective studies could provide crucial insights into the pathophysiology of MC. Such investigations may pave the way for precision medicine approaches, enabling targeted interventions based on microbiome profiles, immune pathways, or genetic predispositions to prevent or mitigate MC in high-risk individuals [4].

Large-scale data analyses using real-world evidence and electronic health record-based pharmacoepidemiologic studies could help to identify at-risk subgroups effectively. This might include older adults with multiple comorbidities, individuals with specific genetic or immune backgrounds, or patients requiring complex drug regimens. By leveraging big data analytics, future research could move beyond broad associations to generate actionable risk prediction models, ultimately guiding clinicians in safer prescribing practices.

Finally, we join Drs. Tome and Pardi in emphasising that meticulous review of both prescription and over-the-counter medications remains essential for any patient newly diagnosed with MC. Even if definitive causality for a single agent is challenging to establish, clinically guided deprescribing or switching to an alternative agent can lead to meaningful symptom improvement [5].

We again thank Drs. Tome and Pardi for their valuable perspective. Their synthesis of current knowledge has underscored the need for continued collaboration to better understand drug-related underpinnings of MC development. We hope that our study, alongside their editorial, will encourage further critical inquiry and research to optimise management strategies for these patients.