Smith, Lucy K. PhD; Draper, Elizabeth S. PhD, FFPH; Manktelow, Bradley N. PhD; Field, David J. DM, FRCPCH
Preterm birth is a major public health problem in countries such as the United States and the United Kingdom, with significant rises in incidence over the last decade,1,2 and represents the leading cause of infant mortality in these countries.3 Advances in perinatal and neonatal care have led to improvements in survival,4 but morbidity is a severe problem, with high risks of neurological impairment and disability particularly among very preterm births, those born at 32 weeks of gestation or less. There are worrisome wide socioeconomic disparities in the incidence of spontaneous preterm birth, with substantially higher rates among women from more deprived areas.2,5 However, there has been a paucity of research to understand this problem, and there has been a call to examine the reasons for these disparities.1
Research to understand the complex mechanisms underlying very preterm birth has recently focused on the potential link between inflammation and spontaneous very preterm birth6,7 and is one of the six key research areas proposed by the March of Dimes Research Agenda.8 Bacterial infection that spreads to the uterus and amniotic fluid is a potential cause of preterm birth.9 Infection is hypothesized to trigger inflammation and consequently spontaneous preterm labor or preterm prelabor rupture of membranes, and the lower the gestation at delivery, the greater the frequency of infection associated with it. Vaginal infections such as bacterial vaginosis10,11 have been shown to be the most common source of infection related to preterm delivery and are generally asymptomatic. Other forms of infection connected to preterm birth include urinary tract infections, chorioamnionitis, and periodontal disease.12 It has been hypothesized5 that differential incidence of infection may explain some of the wide persistent socioeconomic deprivation “gap” in the incidence of very preterm birth between the socioeconomically most deprived and least deprived women, but research is needed to confirm this. Further work also suggests the potential importance of maternal stress in the inflammation causal pathway, a particularly prevalent factor among socioeconomically deprived women.13
Here we used data for a population of very preterm infants born in a former U.K. health region over a 12-year period. In studying this population, we explore whether there was evidence of a link among rates of infection during pregnancy, socioeconomic deprivation, and very preterm birth.
MATERIALS AND METHODS
We studied the former Trent Health Region in the United Kingdom, a geographically defined population with approximately 55,000 births per year and representing approximately 1 in 12 of U.K. births. The Trent Neonatal Survey14 includes information on all very preterm births (between 22 0/7 and 32 6/7 weeks of gestation, inclusive) surviving to receive neonatal care in the region and has ethics approval from the Trent Multicenter Research Ethics Committee and Patient Information Advisory Group approval to collect data without individual consent. This study represents a large-scale study of very preterm birth in the United Kingdom, where data on gestation at birth are not routinely collected. We included data on births between January 1, 1994, and December 31, 2005, to mothers who resided in the former Trent Health Region at birth. Data were not available on 2.5% of very preterm births to mothers who were resident in the area but who were born and received neonatal care outside the region. Previous data verification checks suggest this would not affect the conclusions presented here.
We collected data on gestation, infection, birth weight, mother’s age, type of onset of labor, rupture of membranes, year of birth, and post code. A range of information was collected regarding infection and the use of antibiotics during the antenatal period and labor. For the purposes of this study the presence of infection was determined by whether there was a proven fetal or maternal infection (in both cases using standard clinical criteria as recorded in the mother’s notes) with treatment either given or planned. Diagnosis was based on microbiology results or clinical findings indicating an infection. Cases in which some form of prophylaxis was given with no proven infection were not classified as an infection. We chose this measure of infection because we felt it was less vulnerable to systematic bias that might have resulted from, for example, differing local policies on prescribing antibiotics for situations such as prelabor rupture of the membranes without conventional clinical evidence of infection. Such an effect could potentially have had an effect on the relationship between infection and very preterm birth. Spontaneous births were classified as those births where there was a definite spontaneous prelabor rupture of membranes or where birth was preceded by labor and the labor was not induced. All other births were classified as induced births (including both mechanical induction—caesarean delivery—and medical induction, without evidence of premature labor or rupture of membranes). Because infection is thought to be a causal factor in preterm birth, leading to inflammation and spontaneous preterm labor or rupture of membranes, only data on spontaneous births were analyzed, and induced births were excluded.
We defined gestation according to the hierarchy specified by the national Confidential Enquiry into Stillbirths and Deaths in Infancy program: mother certain of her dates (most reliable); early dating scan (less than 20 weeks of gestation); late dating scan (more than 20 weeks gestation); and postnatal examination (least reliable). If the difference between maternal date and early dating scan was more than 7 days, we chose the early dating scan. Analyses were restricted to singleton births because the incidence of multiple births, eg, twins and triplets, may be affected by socioeconomic factors (eg, access to or ability to pay for infertility treatment), and the outcome of multiple births is affected by a number of factors relating directly to multiplicity. Using centile charts,15 neonates were classified as being small for gestational age if their weight was below the tenth centile for their sex and gestation.
Socioeconomic group was measured using the Index of Multiple Deprivation for 200416 developed for the U.K. Government. It is an area-based proxy measure, calculated at the super output area level, areas representing approximately 1,500 residents. The Index of Multiple Deprivation 2004 is a multidomain measure based on census and administrative data for seven domains of deprivation: (Income, Employment, Health Deprivation and Disability, Education Skills and Training, Barriers to Housing and Services, Crime, and the Living Environment). All super output area levels in the region were ranked by their deprivation score and then divided into five groups (quintiles) with equal populations of children aged 0–4 years from the 2001 census (1, least deprived, to 5, most deprived). For confidentiality reasons, the actual number of births was not available at the super output area level but the use of this proxy was assessed at ward level, comparing quintiles of deprivation based on the actual number of births and those based on the census population data of children aged 0–4 years. We found no significant change in the number of wards in each quintile, and super output area levels were expected to follow a similar pattern. Each preterm infant was assigned to a super output area level of residence based on his mother’s post code at the time of birth and allocated to the appropriate deprivation quintile.
We calculated the proportion of spontaneous very preterm births associated with infection by deprivation quintile, gestational age at delivery, birth weight, weight for gestational age, year of birth, and mother’s age. We calculated odds ratios with 95% confidence intervals using binomial regression models and present the unadjusted and adjusted odds ratios for each explanatory variable.
Between 1 January 1994, and 31 December 2005, there were 7,977 singleton very preterm births surviving to receive neonatal care. Of these, 5,022 births were classified as spontaneous onset of labor, and 2,872 births were classified as induced labor (83 births were unclassified due to missing data [1.0%]). For the 5,022 spontaneous preterm births, data were missing on birthweight, mother’s age, deprivation quintile, or sex for 35 births (0.7%) and these were excluded from the analyses.
Recorded evidence of infection was associated with 24% of the 4,987 spontaneous very preterm births (compared with only 6% of induced very preterm births). Table 1 shows the that proportion of spontaneous very preterm births associated with infection increased with increasing deprivation from 21.4% among women from the least deprived areas to 27.3% among women from the most deprived areas. Infection was also significantly higher in the earlier gestation births and those of lower birth weight, although evidently these factors are highly correlated, because earlier gestation neonates will be lighter. However, there was no apparent difference in proven fetal or maternal infection with weight for gestational age. The actual percentage of spontaneous very preterm births that were small for gestational age is very small at 4.5% compared with 33.9% of induced very preterm births, which indicates that there are different causative factors affecting these two types of very preterm birth. Infection rates also increased with mother’s age, with higher rates of infection among women older than 35 years than younger mothers. There was an increase in infection over time, with higher rates of infection in the last 6 years of the study compared with the first 6 years.
The multivariable model fitted indicated that birth weight and weight for gestational age were not significantly related to infection (P=.11 and .99, respectively) after adjusting for gestation, maternal age, year of birth, and deprivation quintile. Those neonates who were extremely preterm (less than 29 weeks) had a significantly increased odds of infection of over 50% compared with births of 29 weeks or more gestation (P<.001). The odds of infection were significantly increased for mothers aged older than 35 years who had a 30% increase in odds compared with those aged 20 to 35 years (P=.035). Births from 2000 onward had a 19% increase in the odds of infection than those born before 2000. Women from more deprived areas had significantly increased odds of infection (P=.006), and this was most evident in the most deprived quintile, where women had a 43% increase in their odds of infection compared with the least deprived quintile. The association between infection and deprivation strengthened slightly after adjusting for the other risk factors. It did not change significantly with differing gestation, maternal age, or year of birth.
This is a large study of spontaneous very preterm birth and infection. Spontaneous very preterm birth in this study was more likely to be associated with infection among very deprived women than those from more affluent areas. This association between infection and deprivation remained after accounting for higher rates of infection among lower gestational age babies and older mothers.
Complete information was only available on very preterm births and not all births. It is possible that infection rates were higher among all women from deprived areas delivering their babies right up until term and that the increased odds of infection among women from deprived areas is not an explanation for the deprivation gap in the rate of premature birth. However, based on previous research and evidence from our study showing that infection rates reduced with increasing gestation and infection was considerably less common among induced births compared with spontaneous births (6% compared with 24%), we believe infection is a possible explanation for the deprivation gap in incidence. If the lack of recording of infection systematically varied with deprivation then this underreporting of infection would have an effect on the relationship between infection and deprivation shown here. However, we believe it is implausible that an infection was more likely to be reported for more deprived women. Furthermore it is possible that women from more deprived areas were less likely to have an infection recorded because of a reportedly lower uptake of antenatal care services. We think it is possible that differences in the rate of infection between the youngest and oldest groups are partially related to differential rates of uptake of antenatal services, with younger women less likely to attend for antenatal care. Although we showed that infection rates have increased over time, this may in part reflect an improved reporting of infection data. We believe any changes in reporting over time are unlikely to systematically vary with deprivation.
Full information were not available on very preterm infants alive at the onset of labor that did not survive to receive neonatal care (9% of very preterm births in the region). It is possible that these babies could show a different relationship between infection and deprivation. The distribution of these births over the deprivation quintiles was extremely similar to those infants that survived to receive neonatal care, but we could obtain no information on rates of infection.
Current theories regarding infection in the genital tract and spontaneous preterm birth are rather more complex than we have been able to measure in our study. Given the large-scale nature of this study, it was only possible to use routine measures for detection of infection. In contrast, recent theories have focused on the role of subclinical infection and the resulting inflammatory response in bringing about spontaneous preterm labor and rupture of the membranes. This is almost certainly the reason why in this study only 25% of the spontaneous births were associated with infections, whereas others have speculated that it is important in the cause of almost all spontaneous very preterm births. Therefore, it might be reasonable to suggest that what we have measured is a separate phenomenon related to deprivation and not in any way causally to preterm birth. However, the rate of infection was much higher in the those births less than 29 weeks gestation, suggesting that the infection we identified was causally important in leading to preterm birth.
Interventions for infection are not straightforward, because the effect of using antibiotics to eradicate infection on subsequent maternal and neonatal outcomes is not consistent. Trials screening women for infections such as bacterial vaginosis and prescribing antibiotics to women with asymptomatic disease are effective in eradicating bacterial vaginosis, but have little effect on preventing preterm birth and the consequent poor neonatal outcomes.17 Other trials to eradicate infection using antibiotics among women with spontaneous preterm labor have not shown improved neonatal outcomes and suggest that antibiotics should not be routinely prescribed to women in preterm labor without evidence of infection.18,19 However for those women with preterm prelabor rupture of membranes, there is evidence that antibiotic use is associated with improved neonatal morbidity and delayed delivery.20,21 Therefore, further research is needed to understand how outcomes can be improved once infection is detected.
We have used data from a large-scale study to examine interrelationships among very preterm birth, infection, and deprivation and suggest that these findings provide evidence for one possible explanation for the significant excess of very preterm births seen among women from more deprived areas. Here we have used a very specific measure of infection that is easy and cheap to reproduce for monitoring future trends. However, further research is needed to identify whether when more subtle techniques are applied to detect the presence of organisms during pregnancy, a higher prevalence among women from more deprived areas is confirmed. Similarly the differing rates of infection found between the youngest and oldest women in our study need further exploration in this way. Therefore, we hope our research will encourage policy makers to target future research to identify why women from more deprived areas are at increased risk of preterm birth. This will allow appropriate interventions to be developed with the aim of reducing spontaneous very preterm birth and consequently infant mortality.
1. Behrman RE, Butler AS, editors, Committee on Understanding Premature Birth and Assuring Healthy Outcomes Preterm birth: causes, consequences, and prevention. Washington (DC): Institute of Medicine; 2006. Available at: www.nap.edu/execsumm_pdf/11622.pdf
. Retrieved May 17, 2007.
2. Smith LK, Draper ES, Manktelow BN, Dorling JS, Field DJ. Socioeconomic inequalities in very preterm birth rates. Arch Dis Child Fetal Neonatal Ed. 2007;92:F11–14.
3. Bryce J, Boschi-Pinto C, Shibuya K, Black RE, WHO Child Health Epidemiology Reference Group. WHO estimates of the causes of death in children. Lancet 2005;365:1147–52.
4. Draper ES, Manktelow B, Field DJ, James D. Prediction of survival for preterm births by weight and gestational age: retrospective population based study. BMJ 1999;319:1093–7.
5. Kramer MS, Goulet L, Lydon J, Seguin L, McNamara H, Dassa C, et al. Socio-economic disparities in preterm birth: causal pathways and mechanisms. Paediatr Perinat Epidemiol 2001;15 suppl:104–23.
6. Romero R, Gomez R, Chaiworapongsa T, Conoscenti G, Kim JC, Kim YM. The role of infection in preterm labour and delivery. Paediatr Perinat Epidemiol 2001;15 suppl:41–56.
7. Romero R, Chaiworapongsa T, Kuivaniemi H, Tromp G. Bacterial vaginosis, the inflammatory response and the risk of preterm birth: a role for genetic epidemiology in the prevention of preterm birth. Am J Obstet Gynecol 2004;190:1509–19.
8. Green NS, Damus K, Simpson JL, Iams J, Reece EA, Hobel CJ, et al. Research agenda for preterm birth: recommendations from the March of Dimes. Am J Obstet Gynecol 2005;193:626–35.
9. Goldenberg RL, Rouse DJ. Prevention of premature birth. N Engl J Med 1998;339:313–20.
10. Svare JA, Schmidt H, Hansen BB, Lose G. Bacterial vaginosis in a cohort of Danish pregnant women: prevalence and relationship with preterm delivery, low birthweight and perinatal infections. BJOG 2006;113:1419–25.
11. Hillier SL, Nugent RP, Eschenbach DA, Krohn MA, Gibbs RS, Martin DH, et al. Association between bacterial vaginosis and preterm delivery of a low-birth-weight infant. The Vaginal Infections and Prematurity Study Group. N Engl J Med 1995;333:1737–42.
12. Goldenberg RL, Hauth JC, Andrews WW. Intrauterine infection and preterm delivery. N Engl J Med 2000;342:1500–7.
13. Wadhwa PD, Culhane JF, Rauh V, Barve SS, Hogan V, Sandman CA, et al. Stress, infection and preterm birth: a biobehavioural perspective. Paediatr Perinat Epidemiol 2001;15 suppl:17–29.
14. Department of Health Sciences, University of Leicester. Trent Neonatal Survey Report 2003. Leicester (UK): University of Leicester; 2003.
15. Freeman JV, Cole TJ, Chinn S, Jones PR, White EM, Preece MA. Cross-sectional stature and weight reference curves for the UK, 1990. Arch Dis Child 1995; 73:17–24.
16. Noble M, Wright G, Dibben C, Smith G, McLennan D, Anttila C, et al. The English indices of deprivation 2004: report to the office of the deputy prime minister. London (UK): Neighbourhood Renewal Unit; 2004.
17. McDonald HM, Brocklehurst P, Gordon A. Antibiotics for treating bacterial vaginosis in pregnancy. Cochrane Database Syst Rev 2007;(1):CD000262.
18. Kenyon SL, Taylor DJ, Tarnow-Mordi W, ORACLE Collaborative Group. Broad-spectrum antibiotics for spontaneous preterm labour: the ORACLE II randomised trial. ORACLE Collaborative Group. Lancet 2001;357:989–94.
19. King J, Flenady V. Antibiotics for preterm labour with intact membranes. Cochrane Database Syst Rev 2000;(2):CD000246.
20. Kenyon SL, Taylor DJ, Tarnow-Mordi W, ORACLE Collaborative Group. Broad-spectrum antibiotics for preterm, prelabour rupture of fetal membranes: the ORACLE I randomised trial. ORACLE Collaborative Group [published erratum appears in Lancet 2001;358:156]. Lancet 2001;357:979–88.
21. Kenyon S, Boulvain M, Neilson J. Antibiotics for preterm rupture of membranes. Cochrane Database Syst Rev 2003;(2):CD001058.