Impact of Sociodemographic Factors and Maternal Depression On Child Health At One Year of Age
Methods Data from the Fragile Families and Child Wellbeing Study (FFCWB; see http://www.fragilefamilies.princeton.edu/) were used. The FFCWB is a longitudinal survey of 5,000 children born in large cities in the U.S. between 1998 and 2000 (Reichman, Teitler, Garfinkel, & McLanahan, 2001). Children were followed from birth to age 9. Parent interviews were conducted at birth, age 12 months, three years, five years and nine years. Interviews comprised questions pertaining to socio-demographic characteristics, parenting, health, child health and wellbeing. For this study, data from the birth and age one interviews were used. Child health was categorized as excellent/very good, good, or fair/poor. A constructed ethnicity/race variable based on mother self-reported race and ethnicity was used. Symptoms of maternal depression were assessed utilizing the Composite International Diagnostic Interview Short Form (CIDI-SF). Women were categorized dichotomously (yes/no) as meeting depression criteria based on a conservative threshold determined by the FFCWB researchers.
Chi square analyses and logistic regressions were performed to investigate sociodemographic factors at birth and the impact on child health. For proper estimation of regression coefficients and standard errors the Complex Samples module in SPSS 19 was used.
Results Significant differences at birth were found among non-Hispanic Whites, non-Hispanic Blacks, and Hispanics on maternal education level, married/cohabitation status, family income, and low birth weight. Chi square analyses revealed racial/ethnic differences in mother-reported child health status at 12 months postpartum. More Latina mothers reported their children’s health status as fair/poor than either Whites or Blacks. Latina mothers also reported the lowest rates of Excellent/Very Good child health.
To assess the impact of factors on child health status logistic regression was performed. The model contained race/ethnicity, maternal education, maternal age, low birth weight, and maternal depression. The full model containing all predictors was statistically significant. The full model explained 12.3% (Cox and Snell R square) and 21.1% (Nagelkerke R square) of the variance in children’s health status, and correctly classified 87.5% of cases. Only ethnicity/race, maternal education, and maternal depression made a unique statistically significant contribution to the model.
Conclusions and Implications The results demonstrate the impact of child and maternal sociodemographic factors as well as maternal depression on children’s health at one year of age. The findings support the SCF-HSD model which emphasizes the impact of micro factors on disparities in health. Implications for approaching child health research with the SCF-HSD approach and practices to reduce disparities across racial/ethnic groups are reported.