In an influential January 30, 2014 Detroit News article entitled “Detroit is Deadliest City for Children,” the author, Karen Bouffard, wrote, “In 2010, Detroit (population about 713,000) and Cleveland (population about 390,000) had the highest infant mortality rates of Big City America: 13.5 deaths for every 1,000 live births — higher than in Panama, Romania and Botswana. The measure includes deaths from all causes in a child’s first 12 months….” [1. Bouffard, Karen. “Detroit is Deadliest City for Children, “ The Detroit News, January 30, 2014]
After 2010, Detroit’s rate dipped a bit to 12.6/1,000 in 2011 but then jumped to 15.0/1,000 in 2012, the latest year for which statistics are available.[2. Source: 1989-1999 Michigan Death Certificate Registries; 1999-2011 Geocoded Michigan Death Certificate Registries; 2012 Michigan Death Certificate Registry., 1989-1999 Michigan Birth Certificate Registries;2000-2012 Geocoded Michigan Birth Certificate Registries., Vital Records and Health Statistics Section, Division for Vital Records and Health Statistics, Michigan Department of Community Health] The grim reality this article highlighted prompted Detroit policy makers to take action, and they joined other efforts underway in the city and at the state level to combat infant mortality. For example, one of these programs is the Detroit Regional Infant Mortality Task Force’s program Sew Up the Safety Net for Women and Children.
This three-part blog series examines one factor, low birth weight (“LBW”), which is closely associated with higher infant mortality rates.[4.Brown, Sally. “Detroit Task Force to Reduce Infant Mortality,” Henry Ford Health System News and Research, October 19, 2011.] We look at information recorded on the infant’s birth certificate in an effort to understand whether there are demographic and socioeconomic characteristics of the mother that can help identify women most at risk of having a LBW infant. We are not examining medical conditions of the mother – that’s beyond our expertise—nor are we including women who gave birth to two or more babies (e.g., twins or triplets) because infants in multiple births have a known high risk of low birth weight.
In this first blog of the series we focus on Detroit exclusively, investigating the associations between the rate of low birth weight and (1) the mother’s age at the birth of the child; (2) her educational attainment; (3) her marital status; (4) her race; (5) her ethnicity; (6) the adequacy of the prenatal care she received; and (7) the distribution of LBW infants within the city of Detroit (by census tract of mother’s residence).
The second blog focuses on the same factors for the metro Detroit region, which for the purpose of this blog we define as Wayne County outside of Detroit, Oakland County, and Macomb County (“Metro Region”). We compare the associations found for that region with those found for Detroit as a way of asking, “Does place matter?”
If we find that place does matter (and it seems to), we ask in the third blog whether there is something else besides place of residence that can contribute to our understanding of differences in low birth weight rates. In particular we investigate the effect of race and place together.
As we conducted these analyses, we found that the results kept raising more questions than we could answer, so we consider this series to be a starting point for deeper investigations. We hope the report motivates readers to offer their insights and findings and suggest other ways of looking at the birth record data or other sources to better understand factors influencing low birth weight. A further blog series will examine correlates of inadequate prenatal care, another factor associated with higher infant mortality.
Source of data
The data for this study came from birth certificate records of babies born in 2010, 2011, and 2012 with the mother’s residence in Wayne, Oakland, or Macomb Counties. No names or home addresses were included in the records, but we were able to identify the census tract of the mother’s residence. Census tracts with fewer than six births during the three year period were suppressed. Low birth weight is defined as less than 2500 grams.[5.Very low birth weight is defined as less than 1500 grams. In this analysis, very low and low are aggregated as “low” birth weight.] In all the analyses, missing data have been eliminated. Of the total number of women with single births (women with multiple births were excluded) in Detroit during this period, 11 records had missing birth weight information and have been eliminated from all the analyses. Individual variables had differing numbers of records with missing data. In cases where the percentage of missing data was 1% or greater of the total records, we note the percentage of records with missing data. We used three years of data to smooth the year-to-year fluctuations.
Over the three-year period from 2010-2012 there were 30,244 single births (“singletons”) to mothers living in Detroit (30,233 with birth weight on the birth certificate). 88% of these 30,233 were of normal birth weight, while 12% (3,478) were of low birth weight. This is an average of 1,192 LBW singleton babies per year from 2010 – 2012.
We looked first at the relationship between mother’s age and the birth weight of the infant. We divided the ages of the mothers into five groups: teens (younger than 20); 20-24; 25-29; 30-34; 35+. As Table 1 shows, 11 to 12% of the age groups through age 34 had LBW babies. At ages 35 and higher, this percentage increased to 15%.
For women giving birth in Detroit during this three-year period, age did not affect the likelihood of having a low birth weight baby until age 35 and above.
Next we turned to the effect of educational attainment on the likelihood of having a low birth weight baby.[6. 1.5% of the records were missing educational attainment.]
Figure 1 illustrates the perhaps surprising finding that women with an eighth grade education or less had the lowest rate of LBW babies. As we discuss further in the blog, this is due at least in part to the characteristics of the women in this educational category, but we leave this until we have looked at other factors.
There were few women in this Detroit group with a Master’s degree or a doctorate or professional degree (data not shown). The larger group of women who held a bachelor’s degree or higher had a low birth weight rate of 9%, two or three percentage points lower than women with less education.
We divided the mother’s marital status into three categories: never married, married, and widowed or divorced. 2% of the Detroit women (509) were divorced or widowed. Because this is such a small group and one with two different marital situations, we have excluded them from the analysis. Table 3 illustrates that married women had a three percentage point lower rate of low birth weight babies than women who had never been married.
Race and Ethnicity
Race on the birth certificate is coded into 17 different categories. We collapsed the 17 categories into three: Black, White, and Other. Of the 30,000+ women, 81% were classified as Black, 8% as White, and 10% as Other. The Other category includes all those women who were not classified as Black or White (3,092). Because 96% of Other Race was Hispanic, we have eliminated them from the analysis in order to not overlap with the Hispanic identity analysis.
There was a gap of five percentage points in percentage of singleton LBW babies between Black (13%) and White (8%) groups. If Blacks had had the White rate of 8% instead of 13%, there would have been more than 1,100 fewer Black infants of low birth weight during the three-year period.
Table 5 and Figure 2 show that Hispanic women had a substantially lower LBW rate than Non-Hispanic women. Now we can start to tease out why women with an eighth grade education or less had a much lower rate of low birth weight than women with higher educational attainment. First, 59% of the 1,180 women with an 8th grade education or less were Hispanic in contrast to the 10% of the total population of 30,244 women who were Hispanic. Second, those women with an 8th grade education or less were more likely to be married (51%) than the total population of women (19%), and marriage also conferred a birth weight advantage on the infant.
Adequacy of prenatal care
The Kessner Index assigns a value of adequate, intermediate, or inadequate to the level of prenatal care a woman received. The Michigan Department of Community Health Division of Vital Records and Health Statistics explains the Index as, “… a classification of prenatal care based on the month of pregnancy in which prenatal care began, the number of prenatal visits and the length of pregnancy (i.e. for shorter pregnancies, fewer prenatal visits constitute adequate care).”
As is evident from Table 6, both adequate and intermediate prenatal care had similar rates of low birth weight, even if not precisely the same percentage, while inadequate prenatal care, which included no prenatal care, was associated with a five to six percentage points higher rate.[7. 7.4% of the group of 30,000+ women had missing data on this variable.] Note that fully 18% of the women had inadequate prenatal care.
Geographical distribution of low birth weight babies in Detroit
Figure 3 displays low birth weight rates (grouped into ranges) for all census tracts in Detroit. Except for areas of Southwest Detroit (which have a high Hispanic population), there does not appear to be any readily-discernible grouping of census tracts by birth weight.
This first blog examined social and demographic correlates of the birth weight of babies born in 2010-2012 to women residing in Detroit. We looked at the associations between birth weight and the mother’s age, her education, marital status, ethnicity, race, and the level of prenatal care she received.
Overall, ethnicity, race, and level of prenatal care had the greatest effect on birth weight. The Hispanic low birth weight rate was six percentage points lower than the rate for Non-Hispanics, and the rate for Whites was five percentage points lower than the rate for Blacks. Adequate and intermediate levels of prenatal care were associated with a five to six point advantage in birth weight rates over inadequate prenatal care.
Education and marital status were less strongly related to birth weight, although being married versus being unmarried conferred a three percentage point advantage, and having a bachelor’s degree or higher was associated with a two percentage point lower rate of low birth weight than having less education. Age made little difference until age 35 when the rate of LBW babies increased by three-to-four percentage points over younger ages.
In the next blog, we look at the same factors in the “Metro Region”; that is, Wayne County outside of Detroit, Oakland County, and Macomb County considered as one region. We contrast the findings for the Metro Region with what we saw for Detroit to ask, “Does place matter?”
Here you can create the content that will be used within the module.