Research paper
Trajectories of maternal depressive symptoms in the early childhood period and family-wide clustering of risk

https://doi.org/10.1016/j.jad.2017.03.016Get rights and content

Highlights

  • Onset and course of maternal depression is examined from child age 2–54 months.

  • Trajectories: low-stable (84%), high-decreasing (9.5%), moderate-increasing (6.5%).

  • Elevated trajectory groups were in the clinical range for depression at 2 months.

  • The moderate-increasing group was associated with significant family-wide risk.

Abstract

Background

Previous research on individual differences in the course of maternal depressive symptoms has yielded inconsistent findings, with significant variation in the number and pattern of trajectories identified. In addition, missing from the literature is a comprehensive examination of predictors and longitudinal consequences of particular depression trajectories.

Method

Participants in this study included a community cohort of 501 women assessed for depression using the Center for Epidemiologic Depression Scale at infant age 2, 18, 36, and 54 months. A multi-informant approach was used to examine predictors and outcomes of trajectory membership.

Results

Using growth mixture modeling, three distinct trajectories emerged: 84% of the sample demonstrated low-stable levels of depressive symptoms, 9.5% had high-decreasing scores, and 6.5% had moderate-increasing scores. While socioeconomic status and marital conflict differentiated the low-stable group from the high-decreasing and moderate-increasing group, neighborhood collective efficacy differentiated the latter two groups. At 54 months, a clustering of family risks was prevalent for the moderate-increasing depression group, including higher marital conflict and household chaos, lower parental positivity, and heightened levels of child psychopathology.

Limitations

Limitations include reliance on self-reports to assess maternal depression and the relatively small sample size of certain trajectory classes.

Conclusions

The onset and course of maternal depression in the early childrearing period is heterogeneous, with distinct subgroups in the population. Comprehensive assessment of individual, family, and neighborhood stressors augments our understanding of the predictors and consequences of trajectory membership over this critical period of child and family adaptation.

Introduction

Depression is one of the most common psychiatric illnesses, and is considered to be the leading cause of disease-related disability worldwide among women (Kessler et al., 2003). Maternal depressive symptoms are common in the early childrearing period, with meta-analytic estimates suggesting that 19.2% of women experience a depressive episode within 3 months of childbirth (Gavin et al., 2005). Exposure to maternal depression during the first few years of life may be particularly problematic for children, as they are undergoing a period of rapid brain and biobehavioral organization (Shonkoff et al., 2012, Teicher et al., 2003). Indeed, increased severity and chronicity of maternal depression has been shown to be an especially potent predictor of deleterious child outcomes (Ashman et al., 2008, Campbell et al., 2007).

A growing body of research suggests that the course and severity of maternal depression is heterogeneous. In a comprehensive review of the literature, Vliegen et al. (2014) determined that the number of subgroups of maternal depression likely depends on how depression was defined. When depression was measured diagnostically, a group of chronically depressed and remitting mothers most commonly characterized the literature. However, when the severity of depression was taken into account, a third group of mothers with stable-minor depression was identified. These results suggest that there are both qualitative (i.e., course) and quantitative (i.e., severity) differences to consider when examining trajectories of maternal depression.

At present, there is considerable variability in the number and pattern of maternal depression trajectories identified over the early childhood period (Ashman et al., 2008, Campbell et al., 2007, Cents et al., 2013, van der Waerden et al., 2015). These inconsistent findings create difficulty in planning prevention and intervention efforts and targeting sub-groups at greatest risk of psychosocial maladaptation. Although cross-study differences may be attributed to sampling and measurement factors, one critical difference among existing trajectory studies is the methodology used to derive the trajectories. Two methods are often used. The first is latent class growth analysis (LCGA), which assumes that individual differences in symptom course are entirely accounted for by qualitative differences – that is, they are accounted for by distinct trajectories and there is no variability in severity among individuals within each trajectory. Because this method assumes that individual differences are entirely accounted for by qualitative differences, it has a tendency to extract a relatively larger number of trajectories. To date, studies using LCGA have generally identified 4–6 depression trajectories (e.g., Campbell et al., 2007; Cents et al., 2013; van der Waerden et al., 2015). The second method is growth mixture modeling (GMM), which assumes that individual differences in symptom course are accounted for by both distinct trajectories (i.e., qualitative differences between classes) as well as variability among individuals within each trajectory (i.e., quantitative differences in severity within each class). GMM is particularly appropriate for describing the course of depressive symptoms as it accounts for both symptom course and severity (Vliegen et al., 2014). Despite its advantages, only one study to date has used GMM to describe maternal depression across early childhood (Ashman et al., 2008). This study identified three trajectories: 30% of the sample experienced decreasing symptoms, 8% chronic symptoms, and 62% stable-mild symptoms. However, the sample size for this study was small, and non-depressed participants were not included in the analyses. As a result, it is unclear if the number and pattern of trajectories is generalizable to community samples. Moreover, studies examining risk and protective factors of trajectory membership have commonly examined whether membership in the elevated trajectory groups is associated with greater risk for child psychopathology. In the current study, we extend this literature by comprehensively examining individual and family-wide downstream consequences of trajectory membership, including marital conflict, caregiving behavior, family chaos, and child psychopathology.

Section snippets

Putative psychosocial stressors associated with depression trajectories

Etiological models of depression often focus on eventful changes that occurs as a function of stress or life events (Yim et al., 2015). It has been suggested that changes associated with the birth of a child explicitly pull for individual and family re-adjustment, which can intensify psychological stress and create vulnerability for depression or exacerbate pre-existing difficulties (O'Hara et al., 1982, Pearlin et al., 1981). Several antecedent psychosocial stressors have been shown to predict

The current study

The current study examined trajectories of maternal depressive symptoms in a large, community-based sample of mothers assessed at four times over the first five years following the birth of their child. The goals were to: (1) identify heterogeneous trajectories of depressive symptoms, wherein we hypothesized the number and nature of the trajectories would coincide with previous research reflecting chronic, remitting, and low-stable groups (Ashman et al., 2008, Vliegen et al., 2014); (2) examine

Participants

Multiparous women giving birth to infants in the cities of Toronto and Hamilton between 2006 and 2008, who had been contacted by the Healthy Babies Healthy Children (HBHC) public health program (run by Public Health Units), were considered for participation. Inclusion criteria were: (1) English-speaking mother; (2) a newborn weighing ≥1500 g; (3) one or more children less than 4 years old in the home; and (4) agreement to the collection of observational and biological data. Thirty-four percent

Results

Descriptive statistics for all study variables are presented in Table 1. In a preliminary analysis, no significant variations in the linear and quadratic slopes were observed in the GMM models. Constraining these parameters to zero in the GMM 1-class model did not significantly reduce model fit, Δχ2 (2)=2.47, p>.05. For this reason, all subsequent GMMs constrained the linear and quadratic slope variances to zero and only the within-class variation on the intercept was modeled.5

Discussion

The current study examined longitudinal trajectories of maternal depressive symptoms from 2 to 54 months following the birth of their youngest child. Results supported three trajectory classes: 84% of the sample had low-stable levels of depression, 9.5% reported clinically-elevated levels of depressive symptoms at 2 months that remitted by 54 months, and 6.5% showed moderate levels of depression at 2 months that increased up to 54 months. One explanation for the desisting trajectory is that

Study limitations

These results should be considered in the context of several limitations. The first was reliance on maternal reports of depressive symptoms, as symptomatology can vary based on methodological approach (self-report versus clinical interview (Eaton et al., 2000)). However, self-report instruments for assessing depression, considered the most cost-effect methodology, are deemed particularly appropriate when examining levels of depressive symptoms versus the presence or absence of a specific

Acknowledgements

This research was supported by a Banting Postdoctoral Fellowship awarded to the first author. We are grateful to the families who gave so generously of their time, to the Hamilton and Toronto Public Health Units for facilitating recruitment of the sample, and to Mira Boskovic for project management. Sheri Madigan had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. The grant ‘Transactional Processes in

References (49)

  • J.A. Byles et al.

    The gernal funcitoning scale of the family assessment device: reliability and validity

    Fam. Process

    (1988)
  • S.B. Campbell et al.

    Trajectories of maternal depressive symptoms, maternal sensitivity, and children's functioning at school entry

    Dev. Psychol.

    (2007)
  • A. Caspi et al.

    The p factor: one general psychopathology factor in the structure of psychiatric disorders?

    Clin. Psychol. Sci.

    (2014)
  • R.A.M. Cents et al.

    Trajectories of maternal depressive symptoms predict child problem behaviour: the generation R study

    Psychol. Med.

    (2013)
  • S. Cohen et al.

    Stress, social support, and the buffering hypothesis

    Psychol. Bull.

    (1985)
  • J. Coldwell et al.

    Household chaos-links with parenting and child behaviour

    J. Child Psychol. Psychiatry

    (2006)
  • M.J. Cox et al.

    Marital perceptions and interactions across the transition to parenthood

    J. Marriage Fam.

    (1999)
  • P.T. Davies et al.

    Marital conflict and child adjustment: an emotional security hypothesis

    Psychol. Bull.

    (1994)
  • W.W. Eaton et al.

    A comparison of self-report and clinical diagnostic interviews for depression: diagnostic interview schedule and schedules for clinical assessment in neuropsychiatry in the baltimore epidemiologic catchment area follow-up

    Arch. Gen. Psychiatry

    (2000)
  • R. Eisinga et al.

    The reliability of a two-item scale: Pearson, Cronbach, or Spearman-Brown?

    Int. J. Public Health

    (2013)
  • G.W. Evans et al.

    Cumulative risk and child development

    Psychol. Bull.

    (2013)
  • T. Field et al.

    Pregnancy problems, postpartum depression, and early mother–infant interactions

    Dev. Psychol.

    (1985)
  • N.I. Gavin et al.

    Perinatal depression: a systematic review of prevalence and incidence

    Obstet. Gynecol.

    (2005)
  • J.M. Gerard et al.

    In the eyes of the beholder: cognitive appraisals as mediators of the association between interparental conflict and youth maladjustment

    J. Fam. Psychol.

    (2005)
  • Cited by (12)

    • The association between family physical environment and child maltreatment

      2022, Children and Youth Services Review
      Citation Excerpt :

      Moreover, using the U.S. national dataset of Fragile Family and Child Well-Being Study (FFCWS), Jackson and his colleagues revealed that adverse household conditions, in particular health/safety hazards, were significantly associated with child delinquency and self-regulation problems (Jackson et al., 2017; Jackson & Vaughn, 2018). The low-quality family physical environment has also been found to correlate with poor parent-level outcomes such as maternal depression (Madigan et al., 2017; Suglia et al., 2011), anxiety (Suglia et al., 2011), fatigue (Thomas & Spieker, 2016), and poor executive function (Deater-Deckard et al., 2012). For instance, Suglia et al. (2011) found that poor conditions (e.g., inside home dark, crowded, cluttered, dirty, overly noisy) and instability were significantly associated with depression and anxiety among women in fragile families, even after adjusting for other social factors such as intimate partner violence and economic hardship.

    • Intimate partner violence and maternal mental health ten years after a first birth: An Australian prospective cohort study of first-time mothers

      2020, Journal of Affective Disorders
      Citation Excerpt :

      Other studies have identified heterogenous trajectories of maternal depressive symptoms from pregnancy up to 21 years postpartum (Giallo et al., 2014; Mora et al., 2009; Suglia et al., 2011; Skipstein et al., 2012; van der Waerden et al., 2015; Kingsbury et al., 2015). Factors associated with chronic or persistent depressive symptoms include: prior depressive symptoms (predating the index pregnancy) (Giallo et al., 2014; van der Waerden et al., 2015), antidepressant use during pregnancy (Giallo et al., 2014), younger age (Giallo et al., 2014; Skipstein et al., 2012; Kingsbury et al., 2015; Woolhouse et al., 2015), not completing high school (Giallo et al., 2014; Kingsbury et al., 2015), poor relationship quality (Giallo et al., 2014; Madigan et al., 2017; Manuel et al., 2012; Skipstein et al., 2012; Kingsbury et al., 2015), poor physical health (Manuel et al., 2012; Woolhouse et al., 2015), child development problems (Giallo et al., 2014), financial hardship (Manuel et al., 2012; Gavin et al., 2011), housing instability (Suglia et al., 2011), stressful life events (Giallo et al., 2014; van der Waerden et al., 2015; Woolhouse et al., 2015), fewer social networks (Kingsbury et al., 2015) and IPV (Woolhouse et al., 2015; Gavin et al., 2011; Suglia et al., 2011). Notably, only three of these studies included a measure of IPV (Gavin et al., 2011; Suglia et al., 2011; Woolhouse et al., 2015).

    • Maternal adverse childhood experiences, attachment style, and mental health: Pathways of transmission to child behavior problems

      2019, Child Abuse and Neglect
      Citation Excerpt :

      One potential explanation for the association between maternal attachment style and depression with subsequent child behavior problems is that mothers who are depressed and have experienced trauma are more likely to engage in negative and withdrawn parenting behavior (Kluczniok et al., 2016; Lyons-Ruth, Alpern, & Repacholi, 1993; Savage, Tarabulsy, Pearson, Collin-Vézina, & Gagné, 2019). Maternal depressive symptoms have been found to be consistently related to both externalizing and internalizing symptoms in offspring (Downey & Coyne, 1990; Madigan, Wade, Plamondon, & Jenkins, 2017) and hostile-aggressive parenting behavior has been hypothesized to underlie the association between maternal depression and child behavioral problems (Lyons-Ruth et al., 1993). Altogether, mothers with early experiences of maltreatment may be less likely to provide adequate, comfort, guidance, support, and sensitive parenting to their children (Madigan et al., 2006; Zajac, Raby, & Dozier, 2019).

    View all citing articles on Scopus
    View full text