Early-Life Antibiotic Exposure Linked to Increased Risk of Childhood IBD

— Jason Hou, MD, leads a discussion with Shirley Cohen-Mekelburg, MD, and Frank Scott, MD

by Greg LaubDirector, Video, MedPage Today December 15, 2022

Exposure to antibiotics before 2 years of age appeared to increase the risk of pediatric inflammatory bowel disease (IBD), according to a Norwegian nationwide study presented at the Advances in Inflammatory Bowel Diseases annual meeting.

In this first of four exclusive episodes, MedPage Today brought together three expert leaders in the field — moderator Jason K. Hou, MDof Baylor College of Medicine, is joined by Shirley Cohen-Mekelburg, MDof the University of Michigan, and Frank I. Scott, MDof the University of Colorado, Anschutz Medical Campus — for a virtual roundtable discussion on these study findings.

Following is a transcript of their remarks:

New: Hello, everybody. My name is Dr. Jason Hou, associate professor of medicine at Baylor College of Medicine. I’d like to welcome you all to the MedPage Today virtual roundtable. We’re here discussing posters and presentations from AIBD 2022. With me today for the discussion is Dr. Frank Scott, associate professor of medicine at the University of Colorado, as well as Dr. Shirley Cohen-Mekelburg, assistant professor of medicine at the University of Michigan and Ann Arbor VA. Welcome, Dr. Scott. Welcome, Dr. Cohen-Mekelburg.

Scott: Thanks for having me, Jason. Really appreciate it. Looking forward to the conversation.

Cohen-Mekelburg: Same.

New: Great. Well, for today’s discussion, we’ll have several posters that I wanted to go over. Selected out of a lot of different really exciting and interesting topics. But just to pick a couple that I think were particularly salient and relevant to clinical management, clinical care, that we’re dealing with now, as well as stuff that’s likely coming up in the future.

The first poster I wanted to bring up for conversation discussion is an interesting poster about early antibiotic exposure and the risk of developing inflammatory bowel disease. This was a poster by Dr. [Svend] Andersen and colleagues out of Norway titled “Antibiotic Exposure Before Two Years of Age Is Associated With Increased Risk of Pediatric IBD in a Norwegian Nationwide Cohort.” This was an interesting registry-based study. Again, as the title implies, looking at the association of antibiotic exposure, both during pregnancy as well as in early life, 0 to 2 years of age, and subsequent development and diagnosis of inflammatory bowel disease.

In this registry, they identified 754 pediatric IBD patients in the population of about almost half a million children and adolescents. They looked at various risk factors, again, focused on exposures related to pregnancy. And in this association, in their findings, they did find that there appears to be an association related to in utero exposure with mother being on antibiotics during pregnancy and the development of inflammatory bowel disease. Especially when the exposure was in the last trimester of pregnancy, adjusted odds ratio of 1.24, as well as an association for the patient themself when they’re exposed at an early age, before the age of 2, with development of eventual diagnosis of IBD. This is obviously a hot topic related to environmental exposures and risk factors of IBD, since we don’t really know what causes it right now. I wanted to get your thoughts, Dr. Scott, Dr. Cohen-Mekelburg.

Cohen-Mekelburg: I’m happy to begin. So, this population-based registry really provides a unique opportunity to explore environmental exposures and to identify exposures that could potentially drive IBD. I think this is an example of one key priority in IBD research now, which is really focusing more and more on how we might prevent inflammatory bowel disease now that we have a nice armamentarium of IBD-targeted therapies.

Scott: Yeah, I think this is really interesting research. We’ve seen this before in a few other registries. There is a Swedish study that was published several years ago. This effect has been seen in the GPRD [General Practice Research Database]which is in the United Kingdom, as well as the Health Improvement Network, which is also in the U.K. There’s been a few meta-analyses on this topic.

And we’ve seen this effect in other autoimmune diseases as well. There have been associations between early antibiotic exposure and psoriatic arthritis and juvenile rheumatoid arthritis, which have been interesting. What’s great to see in terms of whether or not this is actually an association versus potential confounding is a dose-response here and a clear window of effect, which I think have been seen in other studies, as well as a potential antibiotic class effect with the broader-spectrum antibiotics being potentially more problematic, which has been seen in other studies as well.

Cohen-Mekelburg: Yes, Frank, I think you make a good point about, is this truly an association versus a causal relationship? And I think that’s where a lot of kind of newer methods are going. I think it’ll be interesting to see once they publish their full manuscript, just the statistical methods they used and how they account for follow-up time and adjusting for confounders. One key thing that I was curious about is how they will differentiate exposure to the infection in early life versus antibiotic exposure.

Scott: That’s a great point, Shirley. I think one of the challenges that I’ve always had with these antibiotic association studies across these different autoimmune disease spaces is whether or not this represents actually a causal relationship, because we know that antibiotic exposure can cause long-lasting perturbations in the microbiome versus whether or not the antibiotic prescription itself was an indication of some predisposition to an autoimmune disease or inflammatory bowel disease. And it’s really hard to tease that out when using large secondary registries like this.

Cohen-Mekelburg: For sure. Even though it’s difficult to really study any other way, it’d be hard to do a prospective interventional study of this. So I think the combination of kind of mixed methods here, like this study provides population-level data. You have other studies that have more survey type of data, and the combination is kind of nice, but it’ll be interesting to see where people move forward with this.

New: That’s great. I think both of you brought up very excellent points. As Dr. Scott mentioned, there’s a bunch of other data, this is not necessarily a brand new concept. But we’re dancing around the same challenges that both of you highlighted with these association studies. As an example, Dr. Cohen-Mekelburg, your recent paper looking at NSAIDs and IBD flares. That’s something I think we’ve all talked about for a long time. A lot of the counsel of patients and doing a little bit more rigorous adjustment and analysis. Maybe the association isn’t quite as clear or may not be there at all. So the devil’s in the details.

I agree with you, it will be very interesting, the full methods coming out on this paper. But, again, further adding data to that, the concept, and I’m sure we’ll continue to grow the level of interest in environmental exposures and risk of IBD. So thank you both for that conversation and discussion.