December 1, 2015

expert reaction to study investigating elective caesarean sections and asthma, obesity and all-cause death by 21 years of age in those born

A study published in the Journal of the American Medical Association has investigated between vaginal birth, planned and unscheduled cesarean delivery, and subsequent health problems or early death of the child, reporting differences in risk of a range of outcomes between the three groups. A Before the Headlines analysis accompanied these comments.

(Commenting on the asthma aspect) Prof. Seif Shaheen, Clinical Professor of Respiratory Epidemiology, Queen Mary University of London (QMUL), said:

“Many epidemiological studies have investigated a link between caesarean section and childhood asthma, but the evidence has been conflicting. This latest data-linkage study from Scotland, which is very large, reports an increased risk of asthma associated with planned C-section delivery, compared with vaginal delivery. I am puzzled why they did not also report results comparing unscheduled C-section delivery versus vaginal delivery. There are two key questions to consider in interpreting these findings:

“1. Is this association likely to be causal?  In my view there are a number of reasons to doubt this interpretation: First, a key potential confounder*, namely maternal body mass index (BMI), was not controlled for in the analyses – women having a C-section had a higher BMI than women delivering vaginally (see Table 1), and a higher maternal BMI in pregnancy is associated with an increased risk of asthma in the offspring. Second, when the analysis was restricted to women with no missing data on potential confounders, planned C-section was no longer associated with an increased risk of hospital admission for asthma. Third, two Swedish studies, which used a more sophisticated methodology to rule out potential confounding, found no evidence of association between planned C-section and asthma in children over five years of age.

“2. Even if the link between planned C-section and asthma is causal, is it important and should pregnant women be concerned?  As the magnitude of increased risk was small, and the proportion of women in this study having a planned C-section was also small (less than 4%), this means that the proportion of asthma in the population that might be attributable to planned C-section (assuming a causal link) will also be small.

“In summary, evidence for a causal association is not strong, these findings should be treated with caution, and pregnant women should be reassured.”

* a confounder is a third factor (in this case maternal BMI) which is associated with both the risk factor (C-section) and the disease of interest (asthma), and can lead to an apparent association between the risk factor and the disease which is spurious.

Prof. Jean Golding, Emeritus Professor of Paediatric and Perinatal Epidemiology, University of Bristol, said:

“It is very important to try to assess whether Caesarean section has any long term effects on the developing child. In this paper, however, it is important to note that the authors were unable to determine the reasons why the mothers in this study were having a C-section. The very slight increases in risk found may be related to these reasons rather than the section itself. Unless the reasons can be identified and taken into account I feel that the findings in this paper should not be used when making a decision as to whether a C-section should be undertaken.”

Prof. Kevin McConway, Professor of Applied Statistics, The Open University, said:

“In statistical terms, the researchers have gone a long way in trying to sort out what is related to what. But the huge size of the study – more than 300,000 births over a period fifteen years – is itself an indication of how difficult it is to make sense of this subject. What makes it difficult is that, happily, the great majority of children do not suffer from the diseases and adverse events that were studied. Many thousands of babies have to be studied before the researchers can observe enough cases of asthma or diabetes to draw conclusions at all. Furthermore, the risks of all these diseases are affected by many other things apart from whether the children were delivered by caesarean section or vaginally. While these researchers used appropriate statistical methods to allow for other factors as far as possible, such corrections cannot be perfect, and in any case they can’t allow for things that they have no data on.

“Dr Black, the lead researcher, can appropriately say that “we are confident in saying the delivery could be playing a role in the development of asthma”. But please note the “could”. Dr Black, rightly, emphasises that they cannot say that planned C-section definitely causes a small increase in the risk of asthma. Maybe it does, or maybe the increase is due to something other than the delivery that could not be allowed for statistically. The research report points out that they could not allow statistically for reasons why a C-section was chosen. Maybe it’s one of the reasons for choosing a C-section that leads to the increased risk of asthma, and not the C-section at all. The same goes for the observation that there was a very small increase in the death rate up to age 21 of those born by planned C-section compared to those born vaginally. That could have something to do with the method of delivery, but there are many other possible explanations.

“Further, these results are all for first births, and it remains possible that things are different for subsequent births.

“We mustn’t forget these uncertainties, particularly as the possible increase in risk is really rather small. This research found that, for 1000 children born vaginally, 34 developed asthma requiring hospital admission. For 1000 born by planned caesarean, 37 developed such asthma. Not a huge difference, even if the method of delivery is the cause. And as the researchers point out clearly, if there’s a medical reason why caesarean delivery is preferred, this small increased risk isn’t going to be enough to overturn that preference.

“For me, a key point in the research paper is its last sentence, “Further investigation is needed to understand whether the observed associations are causal”. But that further investigation is likely also to involve thousands of births and many years of further work.”

Prof. Andrew Shennan, Professor of Obstetrics, King’s College London, said:

“This carefully considered analysis of over 321,287 Scottish maternity records demonstrates a clear associative relationship between elective caesarean section, and risk of asthma (admission and drug use) and offspring death.

“The increased mortality (all causes up to 21 years of age) is around an extra 1 death in every 1,000 people born by elective caesarean, which is a considerable increase in this age group (about a 25% increase in relative risk). The increase in asthma is around 3 extra cases of hospitalisation for asthma in every 1,000 people born by elective caesarean.  Effects on other outcomes were either less clear or were reassuring, including on cancer rates.

“In order to suggest any appropriate action as a result of this study we’d need to know the underlying reasons why elective caesarean might be associated with increased asthma incidence and death, and at this stage we don’t. It could not be explained by the increasing belief that exposure to vaginal micro-organisms is protective, as women in labour who needed an emergency caesarean section, and hence babies that were exposed, did not have a reduced risk. This suggests the findings could be related to another factor that prevents vaginal birth, and therefore it remains uncertain that reducing caesarean sections will reverse this problem.

“The authors rightly conclude more research around this association is required. In the meantime from a precautionary principle approach it would seem prudent to avoid unnecessary caesarean sections, which continue to escalate.”

‘Planned cesarean delivery at term and adverse outcomes in childhood health’ by Mairead Black et al. published in JAMA on Tuesday 1 December 2015. 

Declared interests

Prof. Seif Shaheen: “I have no conflict of interest to declare.”

Prof. Jean Golding: “I have no conflicts of interest.”

Prof. Kevin McConway: “I have no relevant interests to declare.”

Prof. Andrew Shennan: “No relevant conflicts of interest.”