Association of late-onset postpartum depression of mothers with expressive language development during infancy and early childhood: the HBC study

Abstract

Background: While it has been implied that an infant's exposure to maternal postpartum depression (PPD) may be associated with delayed development of expressive language, it remains unclear whether such a delay persists into childhood and whether the onset of PPD onset-early (within 4 weeks after childbirth) vs. late (between 5 and 12 weeks postpartum)-is relevant in this context.

Objective: To examine whether children of mothers with early- or late-onset PPD have reduced expressive language scores during infancy and early childhood (up to 40 months of age).

Methods: This longitudinal, observational study was conducted as a part of the Hamamatsu Birth Cohort for Mothers and Children (HBC Study), a population-representative sample in Japan. A total of 969 neonates and their mothers were included in the analysis.

Exposures: Early- and late-onset PPD was measured using the Edinburgh Postnatal Depression Scale.

Main outcomes and measures: Expressive language development was measured using the Mullen Scales of Early Learning. Six points over time were monitored (10, 14, 18, 24, 32, and 40 months postpartum). The relationship between the exposure variable and any change in expressive language score was evaluated using multiple linear regression analysis and growth curve analysis, both adjusted for covariates.

Results: Results from the adjusted regression analysis showed that children of mothers with late-onset PPD had significantly lower expressive language scores at 18 months of age and beyond, with a score reduction of approximately 0.6 standard deviations from the reference value at 40 months of age (95% CI [-0.888 to -0.265], p < .001). This association was confirmed on growth curve analysis, which revealed a significant, monotonic decline of expressive language development between 10 and 40 months of age among children of mothers with late-onset PPD, but not among children of mothers with early-onset PPD.

Conclusion: Exposure to late-onset PPD may lead to a persistent decline in the rate of expressive language development in offspring during infancy and early childhood, highlighting the significance of monitoring for late-onset PPD to facilitate early detection and intervention.

Keywords: Cohort study; Japan; Language development; Postpartum depression.

Figure 1. Flow-chart of study participants.

Figure 2. Expressive language scores among neonates stratified according to exposure to early- or late-onset PPD.

The mean score in the no PPD group was considered as reference. Deviations from the reference value are expressed relative to the sd of the distribution of scores in the no PPD group. The bars (with error bars indicating 95% confidence intervals) correspond to sample-specific mean scores at 10, 14, 18, 24, 32, and 40 months of age, estimated using linear regression adjusted for maternal history of mood/anxiety disorders, infant sex, birth order, twin birth, birthweight, gestational age at birth (in weeks), duration of breastfeeding (in months), maternal and paternal age at the infant’s birth (in years), and maternal and paternal education level (in years). PPD, postpartum depression; sd, standard deviation.

Figure 3. Time evolution of expressive language scores during infancy according to exposure to maternal PPD.

Mean estimate trends for no PPD (green line), early-onset PPD (purple line), and late-onset PPD (red line) are shown. An interaction term between maternal PPD exposure and infant age (in months), which was statistically significant, was entered into the estimating model. Deviations are expressed relative to the sd of the distribution of normal scores. Mean stands for the normal score. The sample-specific means at 10, 14, 18, 24, 32, and 40 months of age were estimated using growth curve modelling adjusted for maternal history of mood/anxiety disorders, infant sex, birth order, twin birth, birthweight, gestational age at birth (in weeks), duration of breastfeeding (in months), maternal and paternal age at the infant’s birth (in years), and maternal and paternal education level (in years).