Testing the role of the timing and chronicity of maternal depressive symptoms in the associations with child behaviour and development

Background There is debate regarding whether the association between maternal depressive symptoms (MDS) and child outcomes is due to the timing or chronicity of symptoms. Objectives To investigate whether critical periods, sensitive periods, or accumulation models provided the best explanation for the association between MDS and children's behaviour and development. Methods Data on mothers (N = 892) were collected from 1996 to 2015 as part of the Australian Longitudinal Study on Women's Health, a prospective longitudinal epidemiological study. Data on children (N = 978, 2-12 years) were collected in 2016/17 as part of the Mothers and their Children's Health study. Mothers were categorised according to whether they reported MDS (scored >= 10 on the CESD-10) before pregnancy, during pregnancy, or in early childhood. Child outcomes were maternal-rated behaviour problems (Strengths and Difficulties Questionnaire; SDQ) and teacher-rated development (Australian Early Development Census; AEDC). We used a structured life course approach to rigorously test critical period, sensitive period, and accumulation (ie chronicity) theories by comparing the fit of a series of models. Results Most mothers did not report MDS at any time (69.2%), 16.9% reported MDS before pregnancy, 13.2% during pregnancy, and 16.5% in early childhood. High/very high total behaviour problems were reported for 7.0% of children, and developmental vulnerability/risk was reported for 15.9% for social competence and 15.7% for emotional maturity. An accumulation model was the best fit, with each period of MDS associated with an increase of 1.71 points (95% CI 1.26, 2.17) on the SDQ and decreases of 0.31 (95% CI -0.50, -0.12) and 0.29 points (95% CI -0.49, -0.08) on AEDC social competence and emotional maturity, respectively. Conclusions Chronic MDS were associated with poorer child outcomes than MDS at any single time. Sensitive and critical period models were not supported. This suggests chronicity of symptoms may be more important than timing.