There is well-documented evidence on how interpregnancy interval (IPI) is associated with adverse perinatal outcomes and how short and long IPIs are associated with increased risk for preterm birth, low birth weight, and intra-uterine growth restriction. However, the extremes of IPI on infant mortality are less well documented. The current study builds on the existing evidence on IPI to examine if extremes of IPI are associated with infant mortality, and also examines if IPI is associated with both neonatal and post-neonatal mortality after adjusting for several known confounders. Matched birth and death certificate data for Arizona resident infants was drawn for 2003-2007 cohorts. The analysis was restricted to singleton births among resident mothers with a previous live birth (n = 1,466) and a randomly selected cohort of surviving infants during the same time-frame was used as a comparison group (n = 2,000). Logistic regression models were utilized to assess the odds for infant mortality at monthly interpregnancy intervals (<6, 6-11, 12-17, 18-23, 24-59, ≥60), while adjusting for established predictors of infant mortality (i.e., preterm birth, low birth weight, and small for gestational age), and other potential confounders. Unadjusted analysis showed greater clustering at extreme IPIs of <6 months and ≥60 months for infants that died (32%) compared to infants that survived (24.7%). Shorter IPI (i.e., <6 months, 6-11 months, and 12-17 months) compared to 'ideal' IPI (i.e., 18-23 months), were associated with infant mortality even after adjusting for confounders. Short intervals were significantly associated with neonatal, but not post-neonatal deaths. IPI above 23 months were not associated with infant mortality in our analyses. Shorter IPIs (18 months or less) significantly increases the risk for neonatal infant mortality even after controlling for known confounders, and our study adds to the existing evidence on adverse perinatal outcomes. Counseling women of reproductive age on the benefits of spacing pregnancies to at least 18 months addresses one preventable risk for early infant mortality.

This study compares the incidence of low birth weight among mothers enrolled in Arizona's Health Start program to a sample of non-Health Start mothers with similar medical and social risk factors. A quasi-experimental design was used to match Health Start program participants to non-participants on the basis of similar medical and social risk factors. Health Start program data were linked to birth certificate data to create a sample of 5,480 pregnant women. A logistic regression analysis was conducted to predict the likelihood of having a normal birth weight (i.e., 2,500 g or more). The findings indicate that Health Start mothers had twice as better odds of having a normal birth weight than non-Health Start mothers, even after controlling for gestational age, adequacy of prenatal care, mother's history of preterm birth, weight gain during pregnancy, alcohol and cigarette use, mother's age, education and residency. Hispanic women in the program were three times as likely to have a normal birth weight baby when compared to Hispanics who were not in the program and twice as likely as non-participant Whites. And lastly, women in urban settings had better birth outcomes, especially Hispanic women. Evidence suggests that newborn infants of mothers enrolled in the Health Start Program had better birth weight outcomes even after controlling for the effects of possible confounders. However, the program seems to affect Hispanics and non-Hispanic Whites differently; in particular, Hispanics who are in the program demonstrated the best birth outcomes. One possible explanation for the general success of the program could be that program participants reported lower cigarette use during pregnancy. A limitation of this study is that that there could be reporting bias on the part of Health Start participants about their risks to enter into the program, which is difficult to verify.

The VES-13 algorithm is robust to substitution of functional items and can be used to identify very high-risk older adults. Multiple imputation of missing items reduces loss-to-follow-up bias and increases sample size.

No abstract