Statistics from Altmetric.com
Question: Is there evidence of an intrauterine effect of maternal depression or anxiety on child attention problems and might the presence of these symptoms at age three account for any association?
People: Single child born to mothers recruited to the Avon Longitudinal Study of Parents and their Children (ALSPAC, n=3442) and Generation R studies (n=2280).
Setting: Generation R Study: Rotterdam, the Netherlands; recruitment of pregnant women with delivery date April 2002–January 2006. ALSPAC study: South West England; recruitment of pregnant women with delivery date April 1991–December 1992.
Risk factors: Parental symptoms of depression and anxiety during pregnancy (18–20 weeks) and when the child was aged 3 years. Generation R Study: assessed using the depression and anxiety scales of the Brief Symptom Inventory(BSI), which assesses symptoms in the preceding 7 days. ALSPAC study: assessed using the Edinburgh Postnatal Depression Scale (depression) and Crown-Crisp Index (anxiety). Both studies considered as covariates of child gender, ethnicity and birth weight, maternal educational level, maternal smoking and alcohol use, and family income.
Outcomes: Child attention problems. In the Generation R Study this was assessed when the child was 3 years old using maternal reports on the Attention Problems syndrome scale of the Child Behaviour Checklist (CBCL1.5–5, with those scoring above the 93rd centile of a Dutch norm group defined as having problems. In the ALSPAC study this was assessed when the child was 4 years old using primary-caregiver reports of child behavioural and emotional problems on the hyperactivity/inattention subscale of the Strengths and Difficulties Questionnaire (SDQ), with those falling in the ‘abnormal’ category defined as having problems.
Design: Two prospective cohort studies.
Follow-up period: Three to 4 years.
In both cohorts, following adjustment for covariates, maternal depressive symptoms during pregnancy were associated with increased risk of child attention problems (Generation R Study: OR 1.23, 95% CI 1.05 to 1.43; ALSPAC: OR 1.33, 95% CI 1.19 to 1.48). There were similar findings for maternal anxiety in pregnancy (Generation R Study: OR 1.24, 95% CI 1.06 to 1.46; ALSPAC: OR 1.32, 95% CI 1.19 to 1.47). Following additional adjustment for maternal depression and anxiety symptoms when the child was age 3 years, the only significant association remaining was for maternal pregnancy depression symptoms with child attention problems in the ALSPAC study (OR 1.17, 95% CI 1.03 to 1.33). Paternal depression, but not anxiety, symptoms during pregnancy showed a borderline significant association with child attention problems in the ALSPAC cohort (OR 1.11, 95% CI 1.00 to 1.24). The associations for paternal symptoms were significantly less strong than for maternal symptoms in the ALSPAC study (depression p=0.04; anxiety p=0.005).
Maternal depression and anxiety symptoms during pregnancy are associated with risk of attention and emotional problems in the young child. The associations are largely accounted for by maternal symptoms when the child aged 3 years, although this did not fully attenuate the effect in one study.
There is accumulating evidence from large population studies that maternal stress (anxiety and/or depression) during pregnancy is associated with increased risk of behavioural and emotional problems in children.1 A range of possible mechanisms exists but a combination of animal and human research findings has led to a relatively settled view that antenatal exposure to maternal anxiety and depression has at least some direct effect on fetal development.2 This has added urgency to calls for psychological intervention during pregnancy to benefit the mother and the future child.
The study by Van Batenburg-Eddes and colleagues provides a challenge to this conclusion through the use of two large population studies (Avon Longitudinal Study of Parents and their Children ALSPAC and Generation R) and the consideration of a range of potential confounding variables, including depression and anxiety in the father. They conclude that a substantial proportion of the association seen between maternal antenatal stress and adverse child outcomes are due to residual confounding, ongoing exposure to postnatal maternal symptoms and genetic effects, leaving a smaller proportion due to intrauterine mechanisms of transmission. Nevertheless, in both cohorts studied, they find that maternal stress shows a consistently stronger risk effect than paternal symptoms; and that even with a wide range of confounders, an elevated risk appears to remain, albeit smaller in one cohort than the other.
The study raises important methodological challenges, particularly what range of confounding variables is the most appropriate to use. It reminds us of the limits of our knowledge regarding mechanisms. However, the findings do need to be considered alongside the many other human and animal studies when conclusions are drawn. Ultimately, experimental designs, including randomised trials of interventions for maternal stress in pregnancy may be the only way to establish the potential causal nature and true strength of these associations; but waiting for these should not dissuade every effort being made to prevent and treat mental health problems in mothers during pregnancy.
Sources of funding: The Wellcome Trust.
Competing interests None.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.