"Antibiotics linked to eczema risk in children," The Daily Telegraph reports, saying that giving babies the drugs increases their chance of developing the condition by 40%.
However, concerned parents should not be put off allowing their children antibiotics on the strength of this news. The research it is based on cannot prove that antibiotics directly cause eczema and when a child is prescribed antibiotics, it is usually for a very good reason and they can be life-saving.
The headlines are based on a review of evidence of whether giving antibiotics during the first year of a baby’s life (or to their pregnant mother) increases their risk of developing eczema.
Researchers identified 20 observational studies, most of which had examined early life rather than pregnancy exposure to antibiotics. The pooled results of the 10 studies that followed children up over time found that early life exposure was associated with a 40% increase in risk of child eczema.
However, there is the potential that the association could be influenced by other unmeasured (confounding) factors. The review does not report on which confounding factors individual studies took into account, and it is likely that these differed between studies. Potential confounding factors that could be associated with both antibiotic use and childhood eczema include:
Most studies also relied on parent or child reports of eczema and antibiotic use, rather than doctor-confirmed diagnoses or medical records.
Overall, this research tells us that children who have eczema are more likely to have been prescribed antibiotics as a baby. However, it cannot tell us conclusively why this is – whether there is a direct risk from the use of antibiotics or whether the observation is due to other factors that explain the relationship.
The study was carried out by researchers from Guy’s and St Thomas’ Hospital NHS Foundation Trust, University of Nottingham and Aberdeen Royal Infirmary. One of the researchers was reported to have received funding from a UK National Institute of Health Research Clinician Scientist Award.
The study has been accepted for publication in the peer-reviewed British Journal of Dermatology.
The media accurately presents the findings of this study, but does not explain the important possible limitations.
The researchers say that a number of studies have suggested that early life exposure to antibiotics may increase the risk of the child subsequently developing eczema, although evidence to date has been conflicting.
The aim of this study was to conduct a systematic review examining the available literature that has addressed the question of whether giving antibiotics during pregnancy (pre-birth exposure) or during the first year of a baby’s life is associated with their risk of developing eczema.
By nature of the question, it is likely that most of the relevant research would be observational cohort studies. In cases where there is consensus that a mother or child definitely needed antibiotics it would be unethical to randomise them to take antibiotics or not, to look at effects on their likelihood of developing eczema.
For this reason, an inherent limitation of this type of review is that the cohort studies included can be affected by other “confounding” factors (such as other health and lifestyle factors) that could be influencing the relationship between antibiotic use and eczema risk. There are measures that can be taken to reduce the effects of these factors in the analyses, but the cohort studies may differ in which factors they take into account and how well they do this.
The researchers searched three online databases to identify observational studies (cohort studies, case-control studies or cross-sectional studies) published up to March 2012 that had investigated a link between any type of antibiotic exposure (pre-birth or during the first year of life) and eczema risk in the child.
The researchers say that, where available, they noted information on the number of courses of antibiotics received, number of infection episodes, doctor consultations, and other potential confounding factors that studies had adjusted for. They looked at risk estimates comparing at least one course of antibiotics to no antibiotics and pooled results in meta-analysis where possible.
Twenty studies met inclusion criteria, 13 of which were prospective studies and seven were cross-sectional. Most of the studies (16) had looked at early life exposure to antibiotics, three at exposure in the womb, and one looked at both. Only five of the studies assessed the outcome of doctor-confirmed diagnoses of eczema, and five studies looked at medical records to provide evidence of antibiotic exposure. Thirteen studies relied on self-report of exposures and outcomes via questionnaires only.
Pooling the results for all 17 studies that had examined early life exposure to antibiotics, showed an overall 41% increased odds of developing eczema in children with antibiotic exposure (odds ratio 1.41, 95% confidence interval (CI) 1.30 to 1.53).
Pooling the results for only the 10 prospective studies that had examined this outcome (that is, excluding the cross sectional studies) gave a similar 40% increased odds of eczema (odds ratio 1.40, 95% CI 1.19 to 1.64). Restricting analyses to only those studies with the highest quality also gave a similar risk figure (OR 1.42).
There was a dose-response relationship, that is, greater dose leading to greater increase in risk. Each additional antibiotic course given during the first year of life was associated with a 7% increase in the odds of eczema (odds ratio 1.07, 95% CI 1.02 to 1.11).
The four studies that had assessed antenatal exposure to antibiotics did not find an association with eczema risk.
The researchers conclude that children with eczema are more likely to have been exposed to antibiotics during the first year of life. There was, however, no link with pre-birth exposure.
This review finds overall that antibiotic exposure during the first year of a child’s life is associated with around a 40% risk of them developing eczema. However, there are important limitations.
All of the studies included were observational rather than randomised controlled trials, and as such there may be confounding factors (other health and environmental factors) that are associated with both the child using antibiotics and with their having increased risk of eczema that are influencing the relationship seen. The review does not report extensively on which confounding factors individual studies took into account, and it is likely that they may have differed between the studies. Potential confounding factors include parental allergic illnesses, parental smoking and other environmental factors, and socioeconomic status. The potential for these confounding factors make it difficult to be certain that antibiotics are directly causing the increased risk, and not that it is other factors associated both with higher likelihood of using antibiotics and of having eczema.
Some of the included studies were reported to be cross-sectional. These single-time assessments cannot reliably tell us that antibiotic use definitely preceded eczema development. For example, a person could be using antibiotics because they have eczema that has become infected rather than vice versa. However, the fact that the researchers conducted separate analyses only including the prospective studies, which found broadly similar results, gives us more confidence in the reliability of the 40% risk figure.
As the researchers say, most of the studies relied on self-reports of eczema and of antibiotic use, rather than diagnoses verified by a doctor, or prescriptions verified through medical records which would be more reliable than questionnaire responses only.
The researchers also acknowledge an additional limitation is that decisions on inclusion of the studies and extraction of information was only performed by one person (reliability can be increased by having two people do these tasks independently and comparing their results).
Overall, this research tells us that children who have eczema are more likely to have been prescribed antibiotics as a baby. What it cannot tell us conclusively is why this is – whether there is a direct risk from the use of antibiotics or whether the observation is the result of confounding from other health and environmental factors that are influencing the relationship.
The observation that increasing dose is associated with increasing risk does support the possibility that the antibiotics may be causing the increase in risk. However, it is not possible to draw a definitive conclusion. The review gives a summary of existing research, and points towards the improvements that could be made in future studies looking at this question.