Tuesday January 18 2011
It should not be assumed that antibiotics cause IBD
Giving children antibiotics can increase risk of irritable bowel syndrome and Crohn's disease later in life, the Daily Mail has reported. The newspaper article says that “scientists believe the drugs may encourage harmful bacteria and other organisms to grow in the gut, which trigger the conditions”.
This study looked at the medical records of over 500,000 children in Denmark, and found that children who had been prescribed antibiotics were more likely to develop inflammatory bowel disease (IBD) than those who had not received such prescriptions. IBD is a group of diseases which includes Crohn’s disease, but not (as suggested by the Mail) irritable bowel syndrome (IBS).
While this study has found a relationship between antibiotic use and IBD, it is not possible to say for certain why such a relationship exists. It might be that antibiotics do raise the risk of IBD, or that the infections being treated with them cause or trigger IBD, or that in some cases antibiotics were being used to treat symptoms of undiagnosed IBD that was later identified. These findings are worth further investigation.
It is important to remember that the risk of IBD in children is very low. In this study of more than half a million children, only 117 were diagnosed with the disease, despite almost 85% of the subjects taking at least one course of antibiotics.
Where did the story come from?
The study was carried out by researchers from the Statens Serum Institut in Denmark and funded by Danish Medical Research Council and the Danish Agency for Science, Technology and Innovation. The study was published in the peer-reviewed medical journal Gut.
This study was reported by the Daily Mail, which has confused inflammatory bowel disease (investigated by this study) with irritable bowel syndrome, which is not an inflammatory bowel disease (and was not investigated in this study).
What kind of research was this?
This was a nationwide Danish cohort study looking at whether there was a link between the use of antibiotics and inflammatory bowel disease (IBD) in childhood. The balance of microorganisms in the intestine has been suggested to be important in the development of IBD. As antibiotics can alter this balance, one suggestion is that their use could potentially affect the risk of IBD.
The main limitation of this type of study design is that the groups being compared (in this case, children exposed and unexposed to antibiotics) may differ in ways other than their use of antibiotics. Any such differences might potentially affect the results and therefore obscure the true relationship. Researchers can try to reduce the likelihood of this by taking such factors into account in their analyses.
Limitations of this nature could potentially be avoided by looking at the risk of IBD in children who had participated in randomised controlled trials of antibiotics, although the practical constraints of such studies mean they would not be likely to include the very large number of children that this study had.
What did the research involve?
The researchers looked at the healthcare records of all Danish children born between 1995 and 2003 who were not part of multiple births (e.g. twins or triplets). They obtained information on collections of antibiotic prescriptions, diagnoses of IBD and other factors that could affect results. They then looked at whether children who had received antibiotics were any more or less likely to subsequently develop IBD compared with children who had not received antibiotics.
The researchers drew data from various national registries to locate eligible children, their filled prescriptions and medical history. The researchers identified:
- all prescriptions for systemic antibiotics antibiotics for internal rather than external (topical) use, given between 1995 and 2004
- the type of antibiotic given, and how many different courses of antibiotics were given in the study period
- all recorded diagnoses of IBD, which includes Crohn’s disease and ulcerative colitis. These diagnoses were identified using records of hospitalisations, emergency department visits and outpatient hospital visits.
The researchers also obtained a range of information on factors which could affect results, including gender, birth order (whether the child was born first, second or third), level of urbanisation of the place of birth, birth weight, length of gestation, mother’s age at the child’s birth, educational level of mother in the year preceding the year of birth, and socioeconomic category of father in the year preceding the year of birth.
However, none of these factors were found to be independently associated with the risk of IBD, so they were not taken into account in the main analyses. These only took into account the child’s age and year of the diagnosis.
What were the basic results?
Overall, the researchers collected data on 577,627 children, with an average follow-up time of about 5.5 years. This provided over 3 million years of data in total. Most of the children (84.8%) had received at least one course of antibiotics.
Across both study groups 117 children developed IBD – 50 of these children had Crohn’s disease and 67 had ulcerative colitis. On average, diagnosis of these conditions was first recorded between the ages of three and four years old.
The researchers reported their outcomes using a measure called the "incidence rate ratio" [iRR], which is the relative proportion of people given a new diagnosis in two different groups within a specified period of time. They found that children who had collected an antibiotic prescription were 84% more likely to develop IBD during follow-up than those who did not [iRR 1.84, 95% confidence interval [CI] 1.08 to 3.15].
When looking at the different types of IBD separately, antibiotics were only associated with an increased risk of Crohn’s disease [iRR 3.41] but not ulcerative colitis. The risk of being diagnosed with Crohn’s disease was greater in the first three months after the prescription collection [iRR 4.43], and greater in children who received seven or more courses of antibiotics [iRR 7.32].
How did the researchers interpret the results?
The researchers concluded that their study is the “first prospective study to show a strong association between antibiotic use and [Crohn’s disease] in childhood”. This suggests that antibiotics or the conditions for which they are prescribed (infections) could potentially increase the risk of IBD or trigger the disease in people who are susceptible.
However, they note that as with all studies of this type, it cannot prove that antibiotics or the illnesses they were prescribed to treat cause IBD. They say that a possible explanation might be that the children had been prescribed antibiotics to treat intestinal symptoms caused by undiagnosed Crohn’s disease that would later be identified.
Overall, this large study has suggested a link between antibiotic use and IBD, although it should not be assumed that antibiotic use is necessarily the cause of the condition. There are a number of alternative explanations for the association, such as the possibility that antibiotics had been given to the children to deal with symptoms of Crohn’s disease that had not yet been diagnosed. Further research will be needed to clarify the situation.
The strengths and limitations of this research must also be considered when interpreting its results:
- The large size of this study, its ability to include most of the children of the relevant age group in the entire country and the level of data available on antibiotic prescriptions are all strengths.
- As exposures and outcomes were based on medical records, the reliability of the findings may depend on the accuracy of the records.
- Standard diagnostic assessments of each child were not carried out, therefore some cases of IBD may have been missed and some children may have been misdiagnosed. However, the authors report that the hospital registers used have previously been found to have a high level of validity and completeness in identifying individuals with IBD.
- Although the prescriptions were filled out, not all of the antibiotics may have been taken by the children. However, this would tend to reduce any link between antibiotics and IBD, rather than make it stronger.
- In this type of study, the groups being compared – children exposed and unexposed to antibiotics – may differ in ways other than their use of antibiotics, and these differences may be affecting the results. Although the researchers took into account factors that they thought could affect results (as the causes of IBD are not well understood), it is difficult to know whether all the important factors have been accounted for.
As the authors acknowledge, it is not possible to say whether the link found is due to the antibiotics, the infection that prompted the need for antibiotics or treatment of existing but undiagnosed IBD.