The distribution of covariates and outcomes for the three maternal age cohorts is reported in Table . Comparing all adolescent mothers (12 to 19 years) to older mothers, newborns of older mothers were, on average, 167 grams heavier than newborns of younger mothers (p < 0.001). Although the younger mothers gained 1.36 kg (3 pounds) on average more during pregnancy than older mothers (p < 0.001), this is not likely to be clinically significant. Compared to older mothers, younger mothers were more likely to be nulliparous (72.6% vs. 29.1%), smoke (27.2% vs. 18.2%), of black race (26.7% vs. 13.8%), and on Medicaid (78% vs. 34.8%), WIC (75.1% vs. 36%), or food stamps (25.5% vs. 16.4%). Not surprisingly, fewer younger mothers were married (20.2% vs. 72.3%) and more had inadequate prenatal care (16.8% vs. 7%). Younger mothers were less likely to have an age-appropriate educational level (72.6% vs. 86.8%), and more likely to have a preterm infant (9.8% vs. 7.6%).
Characteristics of Missouri women with singleton live births, 1997–1999, by maternal age.
Overall, infants born of the youngest mothers (12 to 17 years) were 1.69 (CI 1.24, 2.31) times more likely to die during the neonatal period and 2.47 (CI 1.70, 3.59) times more likely during the post-neonatal period compared to older mothers Table illustrates the risk of neonatal mortality and post-neonatal mortality for the adolescent age cohorts compared to older mothers, stratified by race. There is an increased risk of infant (9.7 deaths/1000 live births; OR 1.95, CI 1.42, 2.67), neonatal (5.3 deaths/1000 live births; 1.63, CI 1.02, 2.47) and post-neonatal mortality (4.4 deaths/1000 live births; OR 2.58, CI 1.56, 4.24) for the 12–17 year old non-Hispanic white mothers compared to older mothers (5.0, 3.2 and 1.7 infant, neonatal and post-neonatal deaths/1000 live births respectively). Among non-Hispanic white mothers 18–19 years of age, there is an increased risk of infant (6.7 deaths/1000 live births; OR 1.35, CI 1.02, 1.79) and post-neonatal (3.0 deaths/1000 live births; OR 1.75, CI 1.11, 2.75) mortality, but not neonatal mortality (3.7 deaths/1000 live births; OR 1.14, CI 0.79, 1.64). No association between age and infant, neonatal or post-neonatal mortality was seen for non-Hispanic black mothers. However, the crude rates of infant, neonatal and post-neonatal mortality for non-Hispanic black mothers were approximately twice those for non-Hispanic white mothers across all age groups.
Infant, neonatal and post-neonatal mortality risk by age group, stratified by race
Table includes the results of the logistic regression model. Race, appropriate education level, marital status, parity, smoking during pregnancy, level of prenatal care and participation in Medicaid, WIC, or food stamp programs significantly affected the primary effect of maternal age on infant mortality risk and were included in the final model. Controlling for these variables, the risk for neonatal and post-neonatal mortality was 1.43 (CI 0.98, 2.08) and 1.73 (CI 1.14, 2.64), respectively, for the youngest adolescent mothers. There was no increased risk of neonatal or post-neonatal mortality for older adolescent mothers. Furthermore, adjusting for gestational age did not appreciably change any of the adjusted OR for the primary effect of age on infant mortality. First-order interactions between covariates and age, considered as a block, did not reach significance and were dropped from subsequent analyses.
Logistic regression models: effect of maternal age on infant mortality.
Of particular interest was the differential effect of the poverty construct in the model on neonatal and post-neonatal mortality risks. Poverty was defined by participation in Medicaid, WIC and/or food stamp programs, and the effect of such programs appeared to be strongly protective against neonatal mortality when controlling for maternal age, but a significant risk factor for post-neonatal mortality. Mothers participating in the food stamps program (our "poorest" category) had a reduced risk of neonatal mortality (OR 0.36, CI 0.25, 0.53) compared to mothers who did not participate in any of the poverty assistance programs. However, their risk of post-neonatal mortality (OR 2.29, CI 1.36, 3.86) was significantly higher. Stratifying the post-neonatal deaths by cause of death appears to show a higher percentage of deaths due to accidents or infections for the youngest mothers compared to older mothers in our study population (Table ).
Underlying causes of post-neonatal death by maternal age (n = 180).