Associations Among Maternal Stress and Language and Socioemotional Development in a Low-income Sample

• In Event: The impact of stress and coping on development: Measurement and mechanisms

Abstract

Background: Maternal stress has been linked with negative outcomes in children's socioemotional and language development, particularly among children being raised in socioeconomically disadvantaged circumstances (Merz, Wiltshire, & Noble, 2018; Shonkoff, 2009). However, stressors can come in many forms, and individual differences exist in how people perceive and respond to different stressors. To date, much of the work examining the relations between maternal reports of stress, language skill, and socioemotional functioning have relied on assessments of a single dimension of maternal stress. However, maternal stress can come from many different sources; the ways that mothers respond to these sources may pose different risks for the emergence of deficits in language and socioemotional functioning.

Methods: Data are from the age 1 (M=13.14 months, SD=2.11) follow-up from 548 mother-infant dyads from four metropolitan areas in the United States who served as the control group for the Baby’s First Years study. The Baby’s First Years study is a randomized control trial of poverty reduction. During each interview, mothers completed a series of questionnaires regarding the different types of stresses they may have experienced. Mothers also reported on their child’s language and socioemotional development (all data are cross-sectional). Maternal reports of stress included perceived stress over the last month (Cohen, Kamarck, & Mermelstein, 1994) and chaos in the home (Evans et al., 2005). Mothers also reported on perceived economic stress over the previous three months and neighborhood safety, since both material hardship and living environment have been linked to stress (Ranjit et al., 2005, Karb, Elliott, Dowd, & Morenoff, 2012). To ensure stress dimensions were not collinear, VIFs were examined and deemed acceptable (all VIFs<2). Socioemotional development was assessed using the Problems Scale from the Brief Infant-Toddler Social and Emotional Assessment (Briggs-Gowan et al., 2004) and language development was assessed using the language subscale of the Ages and Stages Questionnaire (Squires, Bricker, & Twombly, 2009). Covariates included maternal education, race, ethnicity, infant age, site of data collection, and infant sex. Data were analyzed using a series of two linear regressions where all measures of maternal perceived stress and covariates were regressed onto language and socioemotional development.

Results: For socioemotional development, results found that higher maternal perceived stress (B=.192, p<.001, ηp2=.031) and higher household chaos (B=.586, p<.001, ηp2=.071) were statistically associated with more socioemotional problems, while economic stress (B=.078, p=.614, ηp2=.001) and neighborhood safety (B=-.104, p=.670, ηp2=.001) were unrelated to children’s socioemotional development.
For language development, results found that higher household chaos (B=-.373, p=.055, ηp2=.007) and lower neighborhood safety (B=.744, p=.064, ηp2=.006) were marginally associated with worse language development, while economic stress (B=.086, p=.783, ηp2<.001) and maternal perceived stress (B=-.067, p=.525, ηp2=.001) were not statistically related to language development.

Discussion: Together, these findings suggest that different types of stress predict socioemotional and language development for children from low socioeconomic circumstances. Specifically, for socioemotional development, increased maternal report of perceived stress and household chaos show significant unfavorable associations. In contrast, for language development, the data suggest that household chaos and neighborhood safety may be of slight importance.

Authors

• Sonya V. Troller-Renfree, Teachers College, Columbia University
  Presenting Author

• Molly Costanzo, University of Wisconsin - Madison
  Non-Presenting Author

• Katherine Magnuson, University of Wisconsin - Madison
  Non-Presenting Author

• Lisa Gennetian, Duke University
  Non-Presenting Author

• Hirokazu Yoshikawa, New York University
  Non-Presenting Author

• Greg J Duncan, University of California, Irvine
  Non-Presenting Author

• Nathan A Fox, University of Maryland - College Park
  Non-Presenting Author

• Kimberly G Noble, Teachers College, Columbia University
  Non-Presenting Author