Editor’s Choice

The GDM spectrum of disease

There is increasing evidence for the benefits of screening for and treating mild gestational diabetes mellitus (GDM). The HAPO (Crowther et al. N Engl J Med 2005;352:2477–86) and ACHOIS (The HAPO Study Cooperative Research Group, N Engl J Med 2008;358:1991–2002) trials, and a very recent report from the NICHD (Landon et al. N Engl J Med 2009;361(14):1339–48) suggest that even mild GDM that is untreated may be associated with adverse obstetric outcomes. In this issue Tarrico and co-workers (page 1729) go one step further, and report that even in ‘well-managed’ women with GDM, using the commonly employed methods of glucose monitoring, neonates of GDM women show evidence of metabolic derangement compared with neonates of women without GDM. In a study of 37 unaffected pregnancies and 38 pregnancies in women with GDM they collected blood from the maternal radial artery, and from the umbilical vein and artery. They report no significant differences for maternal respiratory gases, acid–base balance and glucose levels between normal and GDM pregnancies. However, they found significantly lower oxygen saturation and oxygen content, and significantly increased lactate concentrations in the umbilical vein of GDM babies, whereas no differences were found for respiratory gases and acid–base balance in the umbilical artery. Glucose concentrations were greater in GDM babies in both the umbilical vein and artery. Such differences would not be surprising in the neonates of women who had poor diabetes control during their pregnancy. However, the women with GDM in this study appeared to have excellent blood glucose control throughout pregnancy, right up until the time of delivery by caesarean section. The percentage of glucose readings that were not performed when they should have been in each time period was only 2% at 32–36 weeks of gestation, 1% at 36–40 weeks of gestation and 1% 2 hours before caesarean section. The percentage of measurements taken that were out of the acceptable range in each time period ranged between 1 and 3%. The authors conclude that these data support the concept that fetuses from GDM pregnancies following standard clinical management have moderate degrees of hyperglycaemia and hypoxaemia, and therefore deserve intense surveillance at term and during delivery. The metabolic derangements they report have not yet been shown to cause a great increase in adverse neonatal outcomes in women with well-controlled GDM, but add weight to the idea that GDM produces a spectrum of adverse effects that continues right down to the mildest and/or most well-controlled cases.

Epilepsy and pregnancy outcome

In contrast to the consensus about the effects of antiepileptic drugs (AEDs) on fetal development and the effects of tonic clonic seizures on fetal oxygenation, there is considerable uncertainty about whether epilepsy is associated with an increased risk of obstetric complications, such as pre-eclampsia, placental bleeding and preterm birth. If such links have been found it is often unclear whether the complications are related to the epilepsy per se or the use of an AED, or to both. A report in this issue by Borthen et al. (page 1736) of a population-based cohort study from Norway tries to answer this question.

They compared pregnancy outcome in women with a current or past history of epilepsy (0.8% of their population) and 362 302 pregnancies in women without a history of epilepsy. Women with epilepsy had an increased risk of mild pre-eclampsia and delivery before 34 weeks of gestation. AEDs were used in 33.6% of the pregnant women with epilepsy. Compared with women without epilepsy, the women who used AEDs had an increased risk of mild pre-eclampsia (OR 1.8), gestational hypertension (OR 1.5), vaginal bleeding late in pregnancy (OR 1.9) and delivery before 34 weeks of gestation (OR 1.5). There was no increased risk of these complications in women with epilepsy not using AEDs.

The study does not provide all the answers. The definition of ‘women with epilepsy’ in their study was epilepsy present now or in the past, and included women whose epilepsy will have resolved many years before, perhaps because they have forms of epilepsy that disappear after childhood or adolescence. In many of these women the risk of seizure during pregnancy will be miniscule, and it is difficult to see why their past illness should have implications for their current pregnancy. In addition, women who have moderate to severe epilepsy just before or during pregnancy are more likely to be receiving treatment with AEDs, so AED use may merely be a marker for more severe forms of epilepsy. What is clear from this study is that women with epilepsy need more careful observation during pregnancy than was previously thought.

Prolapse and its treatment

It is inevitable that globally, increasing longevity will lead to an increase in pelvic organ prolapse (POP), already an important cause of morbidity amongst elderly women. Knowing the factors that predispose a woman to POP may help public health professionals predict the magnitude of that increase and inform strategies for its prevention. Brækken and co-workers on page 1706 report the results of a one-to-one age- and parity-matched case-control study in which they used validated questionnaires, interviews and clinical examinations, including a joint hypermobility scoring system and vaginal pressure transducer measurements, to investigate risk factors for POP. No differences were found between women who did or did not have POP in relation to postmenopausal status, current smoking, current low-intensity exercise, type of birth, birth weight and joint hypermobility. However, body mass index (BMI), socio-economic status, heavy occupational work, anal sphincter lacerations, pelvic floor muscle strength and endurance were independently related to POP. The increasingly ageing and obese population we have is likely to lead to an increasing number of women with POP.

It appears that ring pessary treatment is not the way to treat such women in the long term. In the largest study of its kind to date Sarma and co-workers (page 1715) report what many have know for a long while, that discontinuation rates are very high. In a retrospective case identification study of all of 273 women fitted with either a Portex ring pessary or Introl bladder neck support device during the period 1992–2002, only 167 were successfully using it at 4 weeks. Subsequently, 56% experienced complications comprising bleeding, extrusion, severe vaginal discharge, pain and constipation, in that order. Twenty-three percent had more than one type of complication, and most had more than one episode. Only 14% continued pessary use at the study end point, and their median duration of use was 7 years (inventory quality ratio, IQR, 6–9 years). Of those who discontinued, the median duration was 1.4 years (IQR 0.5–3.6 years). After pessary removal 44% chose conservative treatment, and 30% chose surgery. Now that these data are available, gynaecologists and specialist nurses have reliable information they can give to their patients who are considering using ring pessaries as an alternative to other methods for controlling POP.

Infection and PTD

The link between preterm delivery (PTD) and intrauterine infection has been clear for decades, but strategies to detect early infection in asymptomatic women, and prevent delivery or its inevitable neonatal consequences, have to date been unsuccessful. The detection of increased cytokine levels [interleukin-6 (IL-6), angiogenin or tumour necrosis factor α (TNF-α), for example] has been one area of investigation. Interleukin-18 (IL-18) is a relatively newly discovered cytokine with effects on T-cell activation, which is involved in both innate and acquired immune responses. It has been shown to be increased with microbial invasion of the amniotic fluid in both preterm and term parturition, and in women with preterm prelabour rupture of membranes. Elevated amniotic fluid IL-18 levels appear to correlate with microbial invasion of the amniotic fluid. Levels in cervical mucus and amniotic fluid have been shown to be higher in women with preterm contractions than in asymptomatic women at term. However, the role of IL-18 in preterm delivery as part of occult intra-amniotic infection in asymptomatic pregnant women is not understood. In the study of Daskalakis and colleagues (page 1743), samples of amniotic fluid taken from women at amniocentesis for genetic screening between 16 and 19 weeks of gestation underwent aerobic and anaerobic bacterial cultures, Ureaplasma urealyticum culture and IL-18 assays. The results were compared between 48 women who delivered at <37 weeks of gestation and in 96 women who delivered at term. The preterm delivery group had significantly higher concentrations of IL-18 compared with controls, and the IL-18 level was also significantly higher in women with positive amniotic fluid cultures. More than 50% of preterm deliveries compared with only 12.5% of term deliveries had microbial invasion of the amniotic cavity. Interestingly, in 16.7% of women who delivered preterm there were higher IL-18 levels in the amniotic fluid compared with women who delivered at term, with a negative amniotic fluid culture.

Although as far as neonatal outcome is concerned 37 weeks is not a clinically significant cut-off for premature delivery, this study indicates that IL-18 may be a suitable marker for early inflammatory changes that may predispose a woman to deliver preterm. Whether it proves to be a predictor of PTD below gestations that are more clinically important (e.g. under 32 or 28 weeks gestation) remains to be established.

The obstetric effects of banning smoking

The Republic of Ireland introduced a comprehensive workplace smoking ban in 2004, and this has already been reported to have had positive health effects on the general population (Goodman et al. Am J Respir Crit Care Med 2007;175:840–5; Allwright et al. BMJ 2005;331:1117). Active maternal smoking during pregnancy is an obvious modifiable risk factor for low birthweight (LBW) and preterm birth. Has the Irish smoking ban improved these outcomes? Kabir and colleagues, on page 1782, in a cross-sectional observational study compared the obstetric outcomes in 7593 and 7648 singleton live births dating from 2003 and 2005, respectively, using data from the Euroking K2 maternity system. Maternal smoking rates dropped from 23.4 to 20.6% during this time, and there was a small increase in smoking cessation before pregnancy. They report a 25% decreased risk of preterm births, but a 43% increased risk of LBW infants, differences that were maintained when specific subgroups were examined. The caesarean section rate was about 25% greater in 2005 than in 2003. The finding of a fall in maternal smoking rate must be gratifying to those who instigated the smoking ban, and the fall in preterm births is not unexpected. The rise in LBW infants is more puzzling, as there is strong evidence linking smoking in pregnancy with higher rates of LBW. Equally puzzling was the finding that between the two study periods there was a significant increase in gestational hypertension. Some of the findings of this study are likely to be to the result of other factors that may be unrelated to the effects of the smoking ban.