Classification and heterogeneity of preterm birth


Correspondence: Professor J.-M. Moutquin, CHUS, Fleurimont, Université de Sherbrooke, 3001, 12e Avenue Nord, Sherbrooke, QC, J1H 5N4, Canada.


Three main conditions explain preterm birth: medically indicated (iatrogenic) preterm birth (25%; 18.7–35.2%), preterm premature rupture of membranes (PPROM) (25%; 7.1–51.2%) and spontaneous (idiopathic) preterm birth (50%; 23.2–64.1%). The majority of multiple pregnancies (10% of all preterm births) are delivered preterm (50% for medical reasons). Although medical indications relate more to feto-maternal conditions, PPROM to infections and idiopathic preterm birth to lifestyle, these risk factors are identified in any category, emphasising that preterm birth has a multifactorial origin. Still, several incidences of preterm birth are not completely explained with a plausible cause for PPROM or spontaneous preterm labour suggesting that other causes have yet to be identified. In addition, preterm birth is associated with unrecognised severe congenital anomalies. Variability within the main categories may be explained by the studied population, ethnic group, social class and preventive interventions towards reducing spontaneous preterm birth where the proportion of medically-indicated preterm birth is increased. Despite being retrospective a classification according to gestational age at birth is important for neonatal prognosis. Preterm birth is stratified into mild preterm (32–36 weeks), very preterm (28–31 weeks) and extremely preterm (<28 weeks) with increasing neonatal mortality and morbidity. Recent studies suggested that infection was mostly responsible for extreme preterm birth, while stress and lifestyle accounted for mild preterm birth, and a mixture of both conditions contributed to very preterm birth.


Being born before 37 weeks gestational age or before 259 days, is defined as preterm birth according to the World Health Organisation1. In this definition, the lower limit is not specified. Preterm birth could be qualified according to birthweight with large variations depending on the studied populations or to gestational ages strata. Preterm birth could also be categorised by its clinical presentation: medically induced, preterm premature rupture of membranes (PPROM) and spontaneous preterm labour leading to preterm delivery. Several aetiologies and/or risk factors have been reported for each of the three categories although none completely explain all preterm births. Recent investigations, more directed to defined plausible biological pathways, may reconcile the apparent heterogeneity of preterm birth.

Description of Prematurity

Gestational age at birth is now recognised as a reference standard related to the outcome and prognosis of the preterm infant, together with birthweight. Mild prematurity refers to 32–36 weeks, which could be further subdivided into mild (32–33 weeks) and moderate (34–36 weeks) preterm birth. Mild prematurity accounts for the great majority of all preterm births (Canada: 85%). Although immediate neonatal outcomes are usually reported to be encouraging, this group contributed significantly to an excessive infant mortality in the post-neonatal period (up to one year of age) from asphyxia related conditions, infection and sudden infant death syndrome2.

Birth at 28–31 weeks' gestation is defined as very preterm and accounts for less than 1% of all deliveries and about 10% of preterm births. Immediate survival is expected with a significant proportion of short to long-term morbidity. Below 28 weeks is regarded as extremely preterm (less than 5% of all preterm births) where early neonatal mortality is high with up to 50% of severe handicaps occurring among survivors born below 26 weeks3. Recent reports described survival rates among extremely low gestational ages (24–25 weeks) according to obstetrical variables at admission4.

Categories of Preterm Birth

Preterm birth results from three clinical conditions: medically indicated (iatrogenic) preterm birth, PPROM and spontaneous (idiopathic) preterm birth5. Medically indicated preterm birth in the absence of PPROM or spontaneous preterm labour occurs in about 25% of all preterm births with variations from 8.7%–35.2% according to reports and studied populations5. Medical indications relate to maternal complications such as severe maternal hypertension, abruptio placentae, or endangered fetal well-being, such as intrauterine growth retardation, or non-reassuring fetal state (‘fetal distress’). There are still rare instances of iatrogenic prematurity, when a woman with incorrect dating is electively induced before 37 weeks (Table 1).

Table 1.  Currently recognised aetiological risk factors associated with clinical presentation of preterm birth.
Medically induced preterm birth
  Pregnancy hypertension and vascular disorder
  Medical acute illness or chronic conditions
  Obstetrical complication
  Antepartum bleeding
  Maternal age > 35 years
  Intrauterine growth restriction
  Unstable fetal condition
  Fetal anomaly
  Multiple pregnancies
Preterm premature rupture of membranes
Uterine distension
Cervical anomalies
Afro-American ethnicity
Disadvantaged population
Spontaneous preterm birth
Previous preterm birth, preterm labour
Low body mass, poor weight gain
Strenuous physical workload, ergonomic factors
Uterine anomalies
Psychosocial stress
Lifestyle, smoking
Drug abuse
Maternal age < 18 years

Preterm premature rupture of membranes, usually followed by preterm delivery, accounts for another 25% of all preterm births (range 7.1%–51.2%). This condition occurs more often in the disadvantaged population, and among Afro-American women6. Infection is usually regarded as the main cause of PPROM although, in some cases, it is preceded by spontaneous preterm labour5 (Table 1).

Spontaneous or ‘idiopathic’ preterm delivery accounts for at least 50% of all preterm deliveries (range 23.2%–64.1%)5,7, being more frequent in the population without any established risk factors, where it represents up to 50%–70% of all preterm deliveries according to studied populations8,9. However, a small study identified an aetiological factor in 96% of the 50 reported cases10. Reported risk factors included personal obstetrical history, social factors and lifestyle (Table 1). Spontaneous preterm birth is preceded by spontaneous preterm labour, which cannot be stopped in 70% to 80% of cases8,11.

Multiple pregnancies, mostly twins, have been identified as a major contributor of preterm deliveries (3% multiple birth rate accounting for up to 18% all preterm births)7. About half of all preterm deliveries, among multiple pregnancies, are related to obstetrical, maternal or fetal complications.

Risk Factors

Numerous descriptive studies have highlighted the aetiological heterogeneity of preterm birth, emphasizing its multifactorial origins6,12. A classical epidemiological review13 ascertained well established risk factors: previous low birthweight or preterm delivery, repeated second trimester abortion, uterine and cervical anomalies, in-vitro fertilization, multiple pregnancy, maternal medical complications, gestational bleeding, abnormal placentation, urogenital infection, Afro-American ethnic origin, low socio-economic status, social isolation, smoking and low body mass index (BMI) before conception. Other factors such as maternal age, parity, infertility, heredity, drug abuse, strenuous physical workload, sexual activities, psychosocial stress or stressful life events, inadequate or no prenatal care, maternal weight gain are still debated5,14.

A prospective survey15, carried out in the Quebec City area investigated 117 variables from various risk categories (demographic, medical, obstetrical, anthropometric, ergonomic, socio-economic, lifestyle, psychosocial profile and life events during current pregnancy). This study was carried out among 101 singleton pregnancies in spontaneous preterm labour and 202 control pregnant women without preterm labour, matched for parity, maternal and gestational ages. Twenty-one risk factors were significantly associated with preterm labour in univariate analyses including anthropometric variables, infections, gestational bleeding, ergonomic factors, past obstetric history and psychosocial risk factors. In a multivariate explanatory model16, adjusted for matching variables and socio-economic status, seven risk factors were significantly associated with preterm labour: BMI < 20, previous preterm labour, previous intrauterine growth restriction (IUGR), standing up more than 2 hours a day, abruptio placentae, urinary tract infection and anxiety–stress in the previous 3 months during current pregnancy (Table 2).

Table 2.  Predictive model of risk factors associated with preterm birth among 101 singleton preterm cases and 202-matched control pregnancies16.
 OR95% CI
  1. OR: Odds ratio.

  2. CI: Confidence intervals.

  3. BMI: Body mass index.

  4. PTL: preterm labour.

  5. LBW: low birthweight < 2500 g.

  6. UTI: urinary tract infection.

BMI < 203.962.21–7.09
Previous PTL3.611.12–11.65
Previous LBW2.841.05–7.71
Standing up > 2 hours3.901.53–9.91
Abruptio placentae5.881.20–28.76
Anxiety 12 weeks2.581.20–5.54

This study substantiates that several risk factors of spontaneous preterm labour cannot be modified (previous IUGR, previous preterm labour) while others may respond to targeted interventions (BMI before conception, screening and appropriate treatment of UTI, avoidance of strenuous working conditions together with initiating appropriate interventions and/or counselling when anxiety and stressful life events occur during pregnancy)15.

In another study17 carried out in the Sherbrooke area covering all preterm births for the region (n= 191), a retrospective assessment of risk factors in all categories of preterm birth showed that medical and obstetrical conditions were observed not only in the medically indicated category (100%) but were also encountered in PPROM (28%) and in spontaneous preterm labour (28%), implying that coexisting medical conditions may have influenced the occurrence of either PPROM or preterm labour. Infection accounted for only 25% of all PPROM cases while it was detected in 16% of cases of spontaneous preterm labour. Social risk factors were not reported in cases of medically induced preterm birth but occurred in 22% of PPROM and 33% of spontaneous preterm birth17 (Table 3). Finally, the number of risk factors for each category showed that for medically induced preterm birth, 20% shared at least two conditions while 29% shared three risk factors. With PPROM, 31% had only one risk factor, while 24% and 20% had two and three risk factors, respectively. In spontaneous preterm birth, 38% were associated with a single risk factor, while 20%, 18% and 13% displayed two, three, and four or more risks factors, respectively. No apparent risk factor was identified in about 12% of cases with PPROM or spontaneous preterm birth (Table 4).

Table 3.  Prevalence of risk factors according to subtypes of preterm births, Sherbrooke, Qc. Canada. 1998–199917.
Risk factorMedical %PPROM %Spontaneous %
  1. Medical: Medically indicated preterm birth.

  2. PPROM: Preterm premature rupture of membranes.

Table 4.  Frequency of risk factors (%) associated with clinical presentation of preterm birth according to subtypes of preterm birth, Sherbrooke 1998–199917.
  1. Medical: Medically indicated preterm birth.

  2. PPROM: preterm premature rupture of membranes.


These observations emphasize that preterm birth, whether medically indicated, due to PPROM or from spontaneous preterm labour, is multifactorial but may share similar underlying aetiological mechanisms. Similar observations were reported by Berkowitz14 where some medical conditions (eg. diabetes, hypertension) were significantly associated with all categories together with a previous history of preterm birth or uterine abnormality, ethnicity and inappropriate prenatal care. Sociodemographics, lifestyle or some obstetrical characteristics differed across categories14.

Thus, the apparent heterogeneity of risk factors across the three categories of preterm birth has yet to be substantiated and it is possible that subcategories of PTB may not represent aetiologically different entities14,18.

Towards Better Understanding of the Aetiology

Descriptive associations of risk factors have been of a tremendous help in attempting to identify high risk pregnancies for preterm deliveries5,7,9. However the strongest risk factor (previous preterm birth) is nonexistent for the majority of all preterm births among primiparas5,11. In addition, a substantial proportion of PPROM and spontaneous preterm birth occur without any apparent cause9. Finally, secondary and tertiary prevention strategies with either selected interventions or comprehensive programmes have demonstrated their ineffectiveness towards improved perinatal outcome5,11.

Recent advances in the pathophysiological processes leading to the disruption of uterine quiescence and cervical changes with or without rupture of the membranes identified a few plausible pathways linking underlying biological and psychosocial factors. These included genetic susceptibility, maternal and/or fetal biological/psychosocial stress, inflammatory/infectious causes and mechanical conditions7.

Circumstantial evidence already supports part of these pathways where epidemiologists, clinicians, psychosocial and basic scientists shared their results taking into account that the host susceptibility (a pregnant woman with her own innate and acquired personality traits) may interact with several agents (acute or chronic exposure risks) and the environment (social and cultural context, gender, poverty etc.)19,20.


Preterm birth is a dramatic event for the infant born too early, causing distress for child and parents while also burdening both parents and society. The ineffectiveness of interventions directed towards known risk factors highlights the lack of understanding of plausible causal pathways. The recent transdisciplinary involvement of methodologists and scientists from various disciplines, including psychosocial scientists, may provide landmark advances in innovative interventions to reduce the preterm birth rate worldwide.