Why should preterm births be rising?

BMJ 2006; 332 doi: http://dx.doi.org/10.1136/bmj.332.7547.924 (Published 20 April 2006) Cite this as: BMJ 2006;332:924
  1. A H Shennan, professor of obstetrics (andrew.shennan{at}kcl.ac.uk),
  2. S Bewley, consultant obstetrican
  1. King's College London School of Medicine, St Thomas' Hospital, London SE1 7EH
  2. King's College London School of Medicine, St Thomas' Hospital, London SE1 7EH

    If a rise is confirmed, the implications are considerable

    Preterm birth is a major contributor to neonatal and infant mortality. Few interventions have improved outcome, and management remains an important challenge in modern obstetrics. A paper in this week's BMJ indicates that preterm delivery rates are increasing, which is a worrying prospect.1

    Preterm deliveries account for fewer than 1 in 10 births but result in 75% of neonatal deaths and most neonatal intensive care admissions.2 Preterm birth has considerable impact on long term future health: 1 in 4 survivors born at less than 25 weeks' gestation have severe mental or physical disability.3 Those born at less than 28 weeks spend 85 times as long in hospital as term babies in the first five years of life, with substantial healthcare costs.4 Even beyond 32 weeks, when “neurologically intact” survival is good, educational and behavioural problems occur in 1 in 3 children at the age of 7. This type of morbidity is far more common than overt disability, and 1 in 4 children born between 32 and 35 weeks require support from non-teaching assistants at school.5

    In this week's issue (p 937) Langhoff-Roos and colleagues show a 22% increase in the crude national pre-term delivery rates (before 37 completed weeks' gestation) in Denmark between 1995 to 2004.1 Assisted conceptions, multiple pregnancies, and elective deliveries increased and were also associated with early birth. However, even in a standardised population of low risk European women (20-40 years) there was a 51% increase in early delivery. This is alarming, and also implies that clinicians have failed to have any impact on reducing rates.

    Acute phase interventions such as tocolytics,6 prophylaxis with antibiotics, anti-inflammatory drugs, cervical cerclage, and progesterone, have not improved outcomes. Antenatal use of maternal corticosteroids to enhance fetal lung maturity is the only treatment that has made an impact on important end points.7 In contrast, neonataologists have made major improvements, so the prognosis for an individual baby has never been better. But the goal posts have only been shifted. Obstetricians are more likely to be confident to deliver early. The total health burden to the population will not change unless the number of preterm deliveries can be reduced. The data from Denmark not only show that preterm deliveries are increasing, they also show that preterm births associated with assisted conception are rising. There has been a fourfold increase in triplet conceptions since 1980 in North America, and fewer than 1 in 5 of these pregnancies are spontaneous conceptions.8

    The possible reasons for the findings of Langhoff-Roos and colleagues are numerous, and difficult to elucidate. Preterm birth is associated with many causes,9 and previous preterm birth is the most significant risk factor. Other maternal characteristics include extremes of maternal weight and parity, smoking, low maternal age (< 20), ethnic origin, and social class.10 Obesity and high maternal age at first delivery are risk factors for pre-eclampsia resulting in induction of pre-term labour or caesarean section. However most preterm deliveries follow spontaneous labour, the causes of which are poorly understood. It is probably related to early maturation of a physiological process involving endocrine alterations, inflammation, and uterine stretch with consequent uterine contractions and cervical shortening. Mechanisms and interactions with sociodemographic factors are also unclear. Genetic susceptibility to infection, which is prevalent and important at early gestations, could have an important role.11

    More recently there has been a trend towards earlier ultrasound for dating and screening, which has allowed more accurate confirmation of “menstrual” dates. Menstrual dates have a wide distribution, and the actual gestational age tends to be overestimated. More than a quarter of babies considered to be preterm are more than a week younger than their scan based gestational age.12 Although the authors from Denmark argue that there has been little change in ultrasound policy, and their findings are unrelated to artefact, analysis of neonatal intensive care admissions or morbidity over the same time would allow confirmation of their findings and knowing whether they are clinically important.

    If the trends are real, and confirmed in other countries, the impact for society is considerable. Other countries need to ensure that mechanisms are in place to detect such trends and assess their impact. Untangling the underlying causative factors may be difficult, but general public health measures to do with smoking, teenage and middle age pregnancy, prevention of sexually transmitted diseases, obesity, and social inequities are a good start.

    Room for medical intervention is limited and will depend on further investigation and evaluation. Gynaecologists should optimise surgical treatment of the cervix to avoid cervical damage, as degree of surgical trauma is related to the risk of preterm birth. Regulation maybe required to further limit multiple embryo transfer. Obstetricians should re-evaluate the risks and benefits of delivering babies earlier. If these findings from Denmark are true, the implications for neonatologists, health economists, teachers, parents, and children themselves are worrying.


    • Competing interests None declared.

    • Research p 937


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.