Feasibility of linking maternal study data to offspring vital records data: The Bogalusa Heart Study

With recent findings on developmental origins of disease and intergenerational health effects, data on the outcome of pregnancy is of wide interest.  Such data is routinely collected by state vital statistics registries, but few studies whose focus is not pregnancy outcome have linked to such data. In this project, we report linking women participants in the Bogalusa Heart Study (BHS) to the Louisiana state-issued birth records of their children.  The BHS is a community-based study of cardiovascular health; surveys of the town’s schoolchildren conducted from 1973-1994 and of adults aged 18-50 between the years of 1997 and 2009; the study includes data on almost 6,000 women.  Birth records available for the proposed analysis were those issued by the State of Louisiana between the years 1982 and 2009.  For records issued from 1982-1989, linkage variables available included maternal last name, soundex code for last name, race, and year of birth.  For birth records from years 1990-2009, a three-stage linkage process was performed, with deterministic record linkage based on maternal social security number (SSN), and probabilistic linkage where SSN data were unavailable. 341 women were linked to a birth record in the first stage, and 2,785 women (5,274 infants) in the second stage; 212 women were matched to birth records in both stages.  Together, the two linkage processes identified a total of 3,049 unduplicated women, representing a successful linkage of 52% of all women ever seen by BHS examiners to at least one infant.  This demonstrates the feasibility of linking a large portion of participants in a cohort study conducted for other purposes to offspring birth data.