Methods
248 young African-American women between the ages of 14 and 21 were recruited through obstetrics clinics at an urban teaching hospital, and 221 followed longitudinally until their infants were four months old. Most women were recipients of Medicaid benefits and giving birth for the first time. Following a baseline interview during pregnancy, women were randomized to receive a doula intervention or to be in a control group that received regular medical/social services and a monetary incentive. Doulas were women from the same communities as the mothers who had been teenage mothers themselves and who had received specialized training to provide prenatal education, labor support, breastfeeding counseling, and supporting the parent-child relationship. Doulas made weekly home visits during pregnancy, attended the birth, worked with the mother postpartum in the hospital, and made weekly home visits to three months postpartum.
Mothers and infants were videotaped at four months postpartum while undressing and dressing their infants and playing with a rattle with their infants. Blinded coders used the Parent Child Observation Guide to code maternal sensitivity and maternal encouragement of learning. At four months, mothers also reported on their depressive symptoms (CES-D), parenting stress (Parenting Stress Inventory), and child rearing attitudes (Adult-Adolescent Parenting Inventory).
Results
Using intent-to-treat analyses, there were no differences between the doula and control groups with respect to maternal depression or parenting stress at four months after birth. Mothers who had received doula intervention showed greater encouragement of learning during interaction with their infants than young mothers in the control group. The effect was particularly large for those mothers under the age of 18. Mothers who received doula intervention also reported more appropriate developmental expectations and less role reversing attitudes than control group mothers, but did not differ with respect to attitudes about corporal punishment or empathy toward children.
Conclusions and Implications
There was no intervention effect on depression or parenting stress, although a limitation of the study is that the timing of the followup came after the window of time in which mothers are most likely to experience postpartum depression and before the toddler age period when mothers are most likely to experience parenting stress.
The impact of the intervention on parenting behavior makes this model one of only a very small number of interventions to have documented effectiveness in improving mother-infant interaction during the first months of life. This is important, because parenting behavior during this period has been shown to be associated with later patterns of mother-child interaction and child development.