EXPERT REACTION: Caesarean effects on brain development

Although an interesting study, I’m not sure of its significance and whether the findings are transferable to humans.

It looks specifically at brain and weight changes noticed after different delivery types in mice.

No matter how they are delivered, mice have seemingly programmed nerve cell death after delivery. The authors quote as many as 50 per cent of mice brain nerve cells die during the first postnatal week.

They have said that as far as they knew this phenomenon had never been previously reported (raising the possibility of an artefact rather than a true finding) and I’m not aware of a similar phenomenon occurring in the human, raising the question of relevance in other species.

Mice delivered by Caesarean section don’t exhibit the sudden decrease in brain cell death after delivery (unlike vaginally delivered mice) although this is transient and does return to normal. Additionally, those delivered by C-section have increased body weight at weaning.

The significance of these findings is uncertain, but the authors conclude (without any evidence that similar processes occur in humans) that these changes may indicate neonatal brain developmental changes that could explain the sleep and attention disturbances noted in human infants delivered by Caesarean section.

There is nothing to suggest at this stage that any of the findings described in this article may be transferable to the human population and no comment was made on whether the authors thought increased weight gain was a positive or negative finding.

Birth is a messy business.

At the time of birth the baby is exposed to acute hypoxia by being compressed into the birth canal. It is also exposed to all sorts of bacteria. Surely one would think the more gentle aseptic conditions of a caeasarian would be beneficial to the baby.

However since the work of Patricia Boxsa from McGill University in Canada we have known that there are differences in brain related outcomes particularly stress sensitivity and there are also other health related outcomes associated with C-section in animal systems.

In this study C-section leads to early failures to protect against an acute cell death process 3 hrs post birth in cell death across most brain regions examined. The authors have focused on the paraventricular nucleus (PVN) presumably because of its links to stress responsivity based on earlier studies.

Total cell number in the PVN is normal at weaning but there is a specific reduction in VP cells.  Strange therefore that VP production was not assessed in the C-sectioned animals. They are saying VP is neuroprotective but it fulfils many roles, a reduction in VP cell number may also diminish connectivity to the pituitary that may prove important in altered offspring behaviour in this model. A reduction in VP will also impair stress responsivity. Strange that both VP production and response to stress in weanlings was not assessed.

The authors have been rigorous with controls using maternal CO2 exposure. This control is nice as Boxsa et al., has attempted to replicate C-section by exposing the mouse uterus to mild CO2 i.e. mild hypoxia to replicate protracted birth and found alterations in stress response previously 

Finally number, duration and frequency of pup calls was normal but amplitude reduced in C-sectioned pups. This is a pretty weak behavioural measure and HPA function and stress based behavioural experiments are required to see if there is any permanent behavioural change in these animals.

The report by Alexandra Castillo-Ruiz and others is so flawed that it has no relevance to humans.

Fundamental methodological details are missing from the paper, so that accurate interpretation of data is impossible.

Known confounding factors are disregarded by the investigators.

There is inaccurate reporting of results. There is selective emphasis on some results, which results in biased interpretations and conclusions.

The authors’ hypothesis is not even stated.

Even overlooking the fundamental deficiencies of the report itself, there are serious limitations in the approach taken by the researchers. The gestational timing of brain development in humans and mice is very different. A mouse born at term has a drastically less mature brain than that of a term-born human infant and vulnerability of the brain varies according to maturation, so the effects of Cesarean versus vaginal birth in this experiment using mice cannot be extrapolated to humans.

Another major flaw results from bias introduced by the researchers themselves in choosing the design of their experiments. Mice delivered by Cesarean were “showing no sign of labour”. The effects they attribute to Cesarean delivery may therefore result from a myriad of different factors that distinguish pregnancies where labour is occurring (as in their vaginal delivery group) from those in which it is not.

The publication illustrates much of what is wrong with contemporary medical research: poorly conceived experimentation using animals that are irrelevant to humans in the context of the biological phenomenon being studied; inadequately described methodology resulting in irreproducibility; and biased and misleading reporting of data.