From 2011 onwards, the Women's Health theme will generate new ultrasound-based algorithms for predicting pregnancy complications, with major outputs including novel biomarkers, appearing between 2012-13.

There is widespread concern about rising rates of caesarean section and the cause of rising rates has been unclear. One of the major contributors to overall caesarean rates is poor progress in labour (dystocia) in a first pregnancy and this is the main or joint indication for the majority of primary caesarean sections.

Repeat caesarean section is the other major single indication for the use of the procedure. Hence, addressing the causes of primary caesarean section for dystocia has the potential to impact profoundly on overall CS rates.

Animal studies had indicated that normal preparation of the uterus for parturition occurred over the whole of gestation. We hypothesised that this may also be the case in women. We obtained data from randomised controlled trials of interventions to prevent preterm labour where cervical length was measured in mid-gestation. We were able to perform a secondary analysis where early pregnancy uterine remodelling and preparation for labour was assessed using ultrasonically measured cervical length in mid-gestation. We were able to show that the rate of CS for dystocia at term was strikingly associated with the length of the cervix in mid-gestation (NEJM 2008;358:1346-1453). This suggests that the factors leading to poor progress in labour may be related to dysfunctional uterine remodelling in mid gestation.

We have also addressed the relationship between maternal age and CS risk in some detail. We performed an analysis of population data from Scotland and demonstrated that the risk of primary intrapartum caesarean section increased as a linear function of maternal age from 16 years old upwards. The proportional increase in risk of CS was the same for a five-year increase in age from 16 through to 44. We interpreted this as indicating a biological relationship between maternal age and poor progress in labour.

Furthermore, we estimated that approximately 40% of the rise in CS rates that occurred in Scotland over a >20 year period could be explained by population trends of increasing maternal age at the time of first childbirth. We used BRC funding to perform in vitro studies of uterine smooth muscle obtained from women at the time of caesarean section. This demonstrated that, as the mother's age increased, the amount of spontaneous contraction in isolated strips of myometrium declined and individual contractions were more likely to exhibit dysfunctional multiphasic (coupled) patterns (PLoS Medicine 2008;5:e144). These findings led to the hypothesis that ageing had a biological effect on the uterus. More specifically, we hypothesised that repetitive stimulation of the uterus by oestrogen and progesterone in the years prior to the first birth may be the cause of the continuum of declining uterine contractility over a woman's reproductive life span.

This hypothesis makes a number of predictions. One was that the age of menarche would be similarly predictive of the need for operative delivery. We obtained data from the ALSPAC cohort study and confirmed that a five-year decrease in the age of menarche had a virtually identical association with operative delivery as a five-year increase in maternal age at first birth (BJOG 2009;116:1613-21). The association disappeared on adjustment for the interval between menarche and first birth. Moreover, the association was independent of obvious confounders, such as maternal obesity and height. We will have further opportunity to test this hypothesis. The POP study is collecting information both on both at menarche and contraceptive history which will allow us to test the menarche association in a separate population. Moreover, if the associations between operative delivery and both maternal age at first birth and age at menarche have an hormonal basis, it may be that the associations differ in relation to the use of different forms of hormonal contraception. We will be able to test this using the POP cohort.