Adverse maternal and perinatal outcomes in adolescent pregnancies: The Global Network’s Maternal Newborn Health Registry study

Background: Adolescent girls between 15 and 19 years give birth to around 16 million babies each year, around 11% of births worldwide. We sought to determine whether adolescent mothers are at higher risk of maternal and perinatal adverse outcomes compared with mothers aged 20-24 years in a prospective, population-based observational study of newborn outcomes in low resource settings. Methods: We undertook a prospective, population-based multi-country research study of all pregnant women in defined geographic areas across 7 sites in six low-middle income countries (Kenya, Zambia, India, Pakistan, Guatemala and Argentina). The study population for this analysis was restricted to women aged 24 years or less, who gave birth to infants of at least 20 weeks’ gestation and 500g or more. We compared adverse pregnancy maternal and perinatal outcomes among pregnant adolescents 15-19 years, <15 years, and adults 20-24 years. Results: A total of 269,273 women were enrolled from January 2010 to December 2013. Of all pregnancies 11.9% (32,097/269,273) were in adolescents 15-19 years, while 0.14% (370/269,273) occurred among girls <15 years. Pregnancy among adolescents 15-19 years ranged from 2% in Pakistan to 26% in Argentina, and adolescent pregnancies <15 year were only observed in sub-Saharan Africa and Latin America. Compared to adults, adolescents did not show increased risk of maternal adverse outcomes. Risks of preterm birth and LBW were significantly higher among both early and older adolescents, with the highest risks observed in the <15 years group. Neonatal and perinatal mortality followed a similar trend in sub-Saharan Africa and Latin America, with the highest risk in early adolescents, although the differences in this age group were not significant. However, in South Asia the risks of neonatal and perinatal death were not different among adolescents 15-19 years compared to adults. Conclusions: This study suggests that pregnancy among adolescents is not associated with worse maternal outcomes, but is associated with worse perinatal outcomes, particularly in younger adolescents. However, this may not be the case in regions like South Asia where there are decreasing rates of adolescent pregnancies, concentrated among older adolescents. The increased risks observed among adolescents seems more likely to be associated with biological immaturity, than with socio-economic factors, inadequate antenatal or delivery care. The MNHR is a prospective, population-based observational study that includes all pregnant women and their outcomes in defined geographic communities (clusters). These clusters with approximately 300 – 500 annual births were established in health districts by 7 research sites in western Kenya (Moi University), Kafue and Chongwe, Zambia (University of Zambia), Thatta, Pakistan (Aga Khan University) Belgaum, India (KLE University), Nagpur, India (Indira Gandhi University), Chimaltenango Guatemala (FANCAP), and Corrientes and Santiago del Estero Argentina (IECS). The MNHR was initiated at each of the study sites between 2009 and 2010 and continues to the present. Registry administrators (RA’s), paid study staff who were usually community health workers, nurses, or midwives, identified pregnant women and generally consented those who were eligible by 20 weeks gestation. All women who were residents of the defined communities were eligible and contacted. The RA’s then obtained basic health information at enrollment, and conducted a follow-up visit at or following delivery to collect pregnancy outcomes and health care provided during delivery. A second follow up visit at or after 42 days was done to collect data on maternal and infant health status. Information on the study outcomes was based on medical record review, and birth attendant and family interviews. All study data were collected, reviewed, and edited by staff at each study site. Data were then transmitted to a central data-coordinating center (RTI International, Durham NC) using a secure process, with additional edits performed centrally and addressed at each site. The MNHR study was reviewed and approved by all sites’ ethics review committees (CEMIC, Buenos Aires, Argentina; Francisco Marroquin University, Guatemala; University of Zambia, Zambia; Moi University, Kenya; Aga Khan University; KLE University’s Jawharal Nehru Medical College, Belgaum; Indira Gandhi Medical College, Nagpur), the institutional review boards at each U.S. partner university and the data coordinating center (RTI). All women provided informed consent for data collection and follow-up visits. A detailed description of the MNHR methods can be found elsewhere [15]. The study population for this analysis was restricted to women enrolled in the MNHR, aged 24 years or less, who gave birth to infants of at least 20 weeks’ gestation and weighing 500g or more. The study period included women enrolled with deliveries January 2010 through December, 2013. The exposure of interest was adolescent maternal age at enrollment categorized into two groups: <15 years (early adolescence), and 15-19 years (older adolescence). The World Health Organization (WHO) defines adolescents as those aged 10 to 19 years [3]. Mothers in the age category 20-24 years were the reference group. As the aim of this study was explanatory, the selection of confounders was based on a conceptual hierarchical framework oriented to distinguish potential confounding factors from mediating factors [16]. Maternal education and parity are distal socioeconomic and reproductive factors of adverse pregnancy outcomes that are also associated with adolescent pregnancy [3]. Although data on family income are not collected in the MNHR, the clusters are located in low resource settings in which the vast majority of women are of low socioeconomic condition. Maternal height and pre-pregnancy weight in the Registry had differential missing rates that did not permit these to be included in this analysis. Low pre-pregnancy maternal body mass index (BMI) is associated with preterm birth and LBW, and adolescents tend to have lower BMIs than their adult counterparts [13]. However, low BMI may also be an indicator of biological immaturity in adolescent girls; thus if adolescence was a risk factor of adverse pregnancy outcomes, BMI might be more a mediator in the causal pathway than a confounder, and would not be used for adjustment. Similarly, adolescents may have a different access to, and quality of antenatal and delivery care than the adult mothers. An adverse outcome could be, partially at least, mediated by a lower access to care or lower quality of care. Thus the comparison of the antenatal and delivery care processes between adolescents and adults would be used for the interpretation of the mechanism of action of maternal age on adverse outcomes, rather than as potential confounders. Therefore, to control for confounding, if parity and education were clinically different among the maternal age groups in the univariate analysis we would adjust for these factors in the multivariate analysis. Figure ​Figure11 shows a simplified conceptual hierarchical model of the relationships between age and other factors with adverse pregnancy outcomes, adapted from Victora et al [16]. Simplified Conceptual hierarchical framework We considered the following maternal outcomes: antepartum and postpartum hemorrhage, obstructed labor, hypertensive disorders, maternal sepsis, and maternal mortality at 42 days postpartum. The perinatal outcomes were: preterm birth (live birth at <37 weeks’ gestation), LBW (live birth weighing <2,500g at birth), stillbirth (fetal deaths occurring >500 g [or >22 weeks gestation]), early neonatal deaths (neonatal deaths 0-6 days after birth), neonatal deaths (neonatal deaths 0-28 days after birth), perinatal deaths (neonatal deaths 0-6 days plus stillbirths). Descriptive analyses included calculating the frequency and distribution of values. We compared the frequency of maternal characteristics and the process of antenatal and delivery care between the adolescent groups and adults. The interpretation of the differences was done on clinical basis, acknowledging that with these large sample sizes, small but clinically not relevant differences would be statistically significant. To estimate the effect of the adolescent age categories on maternal and perinatal outcomes, generalized linear models were used evaluate the relationship of adolescent age and adverse pregnancy outcomes and to develop point and interval estimates of the relative risk associated with these risk factors; generalized estimating equations were used to account for the correlation of outcomes within cluster in developing appropriate p-values and confidence intervals. All data were analyzed using SAS v.9.3 (Cary, NC).