Role of Maternal Depression on Child Development: A Prospective Analysis from Pregnancy to Early Childhood

Abstract

Few studies have examined how different characteristics of maternal depression may be associated with developmental outcomes among low-income children. The current study prospectively examined whether the timing (pregnancy vs. early postpartum), severity, and chronicity of maternal depression were associated with child cognitive and social-emotional development in two cohorts of primarily low-income Latinx immigrant mothers and their children. Maternal depression was assessed during pregnancy and at 6 months postpartum. Child development was assessed up to 5 years postpartum. Results showed that maternal depression experienced during pregnancy was associated with lower child cognitive development, particularly among girls. Additionally, both the timing (pregnancy and early postpartum) and severity/chronicity of maternal depression were each independently associated with lower child social-emotional development. These findings highlight the need for early prevention interventions to help offset the adverse effects of maternal depression on child developmental outcomes in this at-risk population.

Introduction

Major depression is one of the most common mental health disorders in the U.S. and is a major public health issue for women [1]. Women are twice as likely as men to develop major depression, with the greatest risk for a first major depressive episode occurring during their childbearing years [1]. In fact, approximately 20% of women experience clinically significant symptoms of depression during pregnancy, with 10% to 15% of affected women continuing to experience depressive symptoms following childbirth [1, 2]. Much higher rates of postpartum depressive symptoms (30% to 60%) have been reported among low-income and ethnic minority mothers [2, 3] with Latina mothers and their children showing substantial increases in depression the longer they stay in the U.S. [4]. These rates of depression are significant in that once women experience their first depressive episode, they are 60% to 80% more likely to experience a recurrent episode in their lifetime [1]. Therefore, depressive symptoms occurring at critical junctures in one’s life, such as when one becomes a mother, could have serious long-term consequences for both the mother and her child [5].

A significant body of research has shown that maternal depression can increase a child’s risk for cognitive and social-emotional problems from infancy throughout adolescence [6, 7]. For example, studies demonstrate that infants of depressed mothers, compared to infants of non-depressed mothers, have worse motor development, less secure attachment with their mothers, and more difficult temperaments [8]. Furthermore, as toddlers, children of depressed mothers tend to have problems with emotion regulation and delays in language development that then persist and lead to poor academic performance, impaired social skills, and a greater prevalence of behavior problems as they begin to enter school [9].

Although there are multiple pathways by which maternal depression may affect child development, one mechanism that may have a lasting influence is the process of fetal programming. Fetal programming refers to the process by which a major event, occurring during critical periods in the development of the fetus in utero, can have a long-lasting or permanent influence on subsequent development during the postpartum period [10]. Given that pregnancy is a time of enormous change and physiological development for the fetus, impactful disturbances (such as maternal depression) can lead to long-lasting effects on the fetus. In particular, maternal depression can affect organ development (e.g., prefrontal cortex), growth rates, and the hypothalamic–pituitary–adrenal axis (HPA axis; the system responsible for regulating stress hormone production, e.g., cortisol) such that the fetus becomes overly sensitive to environmental stressors [9, 11]. Indeed, studies have shown that chronically elevated levels of cortisol in depressed mothers during pregnancy can lead to slower fetal growth, premature birth, infant brain cell damage, and over-reactive infant biological stress responses, which can cause long-term cognitive and emotional deficits later in childhood [8, 11].

Another mechanism by which maternal depression can affect child development, especially during the postpartum period, is the quality of the social interaction that a mother has with her child. Typically, mothers experiencing postpartum depression may be more socially withdrawn, exhibit more negative affect, be less emotionally available, and use under- or overstimulation when interacting with their child. In turn, this may affect their child’s ability to learn how to regulate or express their emotions, learn how to interact with others, and reduce learning opportunities to enhance cognitive and language development [5–8]. These effects may be even more pronounced among children of chronically depressed mothers [9]. Given the long-term impact of these effects, prospective studies are needed that examine how different characteristics of maternal depression lead to developmental problems for children.

Characteristics of Maternal Depression and Child Development

Although it is widely recognized that maternal depression can adversely affect child development, few studies have prospectively examined whether the timing, severity, and chronicity of maternal depression affect developmental outcomes in children. The form and severity of developmental impairments in children of depressed mothers may vary depending on whether depressive symptoms appear during pregnancy or the early postpartum period. Few prospective studies have examined how maternal depression during pregnancy affects later child functioning. In terms of cognitive outcomes, results from two studies show mothers’ prenatal depression to be associated with their children having lower cognitive development at 18 months of age [12] and a lower IQ at 8 years of age [13]. Results on social-emotional functioning (e.g., play behavior, prosocial interactions) have been mixed, with two recent studies demonstrating mothers’ prenatal depression to be associated with their children having lower social-emotional development at 2 [14] and 4 years of age [15]; a third study showed non-significant effects at 1 year of age [16]. In terms of postpartum depression, the few existing studies show mixed results for cognitive development, depending on the domain being assessed. Specifically, depression experienced by mothers during the first six months postpartum have been associated with their children having lower scores on fine motor skills and expressive language after 18 months of age, but no effect on gross motor skills or receptive language [17]. In addition to its effects on cognitive development, maternal depression experienced during the first 12 months postpartum has been found to be associated with lower social-emotional development for children at 2 to 4 years postpartum [14, 15]. Collectively, these findings highlight the need to prospectively examine the timing of maternal depression on later child developmental outcomes.

In addition to the timing of maternal depression, the severity and chronicity of depressive symptoms experienced by mothers have also been associated with child developmental outcomes. In particular, studies have shown mothers who experience more severe depressive symptoms and those who are more chronically depressed to have children with more cognitive impairments and social-emotional problems later in childhood [13, 14, 18]. Moreover, these studies have shown boys of chronically depressed mothers to perform worse on cognitive assessments than girls, which may reflect gender differences in how biological markers of stress and depression (e.g., cortisol) differentially impact neurodevelopment during pregnancy (e.g., more delayed physical and neuromuscular development in boys versus girls; [19, 20]). However, few of these studies have examined whether the timing, severity, and chronicity of maternal depression affect the developmental outcomes of vulnerable populations, such as children of low-income families.

Maternal Depression and Vulnerable Child Populations

In general, studies have shown maternal depression to have a greater negative impact on the developmental outcomes of children from lower income families, with these children exhibiting lower levels of cognitive functioning, academic achievement, and language and social development than children from higher income families [9, 21]. Some studies have suggested that low-income status may negatively impact the development of children through both the presence of chronic maternal depression and stress (which could lead to poor parent–child interactions) and having less access to stimulating learning materials and home environments [22, 23]. Less often studied for developmental problems are children of low-income, immigrant families. Nearly 70% of U.S. immigrants are currently from Mexico or other parts of Latin America. Latinx currently number 60 million or 18% of the total U.S. adult population and 22% of the total U.S. child population [24, 25]. One of the few studies to examine early developmental outcomes among second-generation children of Latinx immigrant families in the U.S. showed that these children had lower vocabulary scores than children of native-born families, particularly if Spanish was spoken at home (notably, all child assessments were conducted in English). However, no differences were observed on non-verbal cognitive domains or behavioral functioning [26]. Chronically elevated levels of cortisol in depressed mothers during pregnancy may particularly impact the language development of second-generation children of Latinx immigrant families by impacting fetal growth rates and development of biological systems related to attention, which have been associated with later language delays [8, 11]. These effects may be exacerbated during the postpartum if the mother’s depression is chronic and interferes with their ability to use play and reading opportunities, or if they do not have the necessary instructional resources or knowledge (e.g., access to language-specific books, educational toys) to extend their children’s language skills [9]. These developmental patterns are particularly salient given the substantial increases in depression the longer Latinx families reside in the U.S. due to exposure to ongoing acculturative stressors of adapting to a new country (e.g., language barriers, isolation from extended family network, discrimination) that leads to increased cross-generational depression risk, with their U.S. born children showing even higher rates [4, 27]. Together, these results demonstrate the need to examine how different characteristics of maternal depression may be associated with the developmental outcomes of low-income and second-generation children of immigrant families over time.

Present Study

The purpose of the current prospective study was to examine whether the timing (pregnancy vs. early postpartum), severity, and chronicity of maternal depression were associated with child cognitive and social-emotional developmental outcomes in low-income mothers and their children (the majority of mothers whom were Latinx immigrants to the U.S.). It was hypothesized that children of mothers who displayed more severe and chronic depression, during pregnancy and the early postpartum, would display lower levels of cognitive and social-emotional development later in childhood. The effects of maternal depression on these developmental outcomes were expected to be more prominent among boys than girls.

Method

Participants

Two cohorts of pregnant women were recruited from a public sector hospital in northern California between 1998 and 2001 (see Fig. 1). This hospital serves a predominantly low-income population representative of the surrounding neighborhoods (80% with public health insurance or uninsured; [28]). Women were recruited either by research staff who approached them in the waiting room during their prenatal clinic visits, through print-based advertising (brochures, flyers), or by referrals from their health care provider. Recruitment efforts were part of the Mothers and Babies: Mood and Health Project, which was designed to identify risk factors and prevent the onset of major depression among low-income, pregnant women [29, 30]. Women who were interested in the study then signed a consent form to determine their study eligibility and to participate in the study. Eligibility criteria included being 18 years of age or older, 6 to 32 weeks pregnant, fluent in either Spanish or English, and free of any major medical or substance use problems (i.e., gestational diabetes, high risk pregnancy).

Fig. 1.

Participant retention rates for Mothers and Babies: Mood & Health Project

Procedures

Eligible women were administered depression assessments [using the Center for Epidemiological Studies—Depression Scale (CES-D), described below] during pregnancy and at 6 months postpartum. Women who had participated in these earlier depression screenings were re-contacted via mail and/or phone approximately 1 to 5 years later to have their child participate in an assessment of their cognitive and social-emotional development. Of the 139 pregnant women who completed depression screenings during pregnancy and at 6 months postpartum, 96 women and their children (69%) were successfully contacted and participated in the follow-up child development assessments at 1½ to 4½ years postpartum. Of the remaining women, 34 were lost to contact and nine declined to participate (see Fig. 1 for retention rates for each of our age cohorts). Most of these women were from the older child cohort (i.e., n = 36; 3 to 4½ year of age) given that 2 years had elapsed since their 6-month postpartum assessment and many of the women had moved or changed their phone number during that time. Participants did not differ from non-participants by demographic characteristics or prenatal depression level (p > 0.05).

Upon agreeing to participate in the follow-up assessment and providing their informed consent, mothers were asked to attend two separate clinic appointments, occurring one week apart, to assess the cognitive and social-emotional development of their child who was between 1 and 5 years of age. During the first clinic visit, a bilingual (Spanish and English-speaking) member of the research staff assessed the child’s cognitive development using the Mullen Scales of Early Learning (see description below) in a private interview room. This research staff member had received extensive training to conduct the cognitive assessments in Spanish or English, which were videotaped, and reviewed by a licensed clinical child psychologist (also bilingual) to validate the scoring of each child’s performance.

During the second clinic visit (approximately one week later), a semi-structured interview with the child’s mother was conducted by another bilingual member of the research staff (also trained and supervised by a licensed clinical child psychologist) using the Vineland Social-Emotional Early Childhood Scales in order to assess their child’s social-emotional development. All assessments (for both clinic visits) were conducted in Spanish or English depending on the mother’s and child’s preference. Children who performed below average on any of the developmental assessments were referred to either their physician (e.g., hearing, visual problems) or to a child clinic (e.g., cognitive delay, social-emotional problems) for further evaluation. Likewise, mothers who were currently depressed were referred for a formal clinical assessment. All study procedures were approved by the Institutional Review Board at the University of California, San Francisco.

Measures

Sociodemographic Questionnaire

The sociodemographic questionnaire assessed for maternal characteristics, such as maternal age (in years), number of weeks pregnant, parity status, total years of education, annual household income, marital status (married/living with a partner, single), primary language spoken at home (Spanish, English), country of birth, and age when they immigrated to the U.S. (in years). Child characteristics were also assessed and included the child’s gender (girl, boy), number of birth complications, the child’s age (in months), the child’s primary language (Spanish, Bilingual, English), whether or not they were breastfed, and number of siblings. Number of birth complications was calculated using a Medical Complication Index, which consisted of adding nine birth complication factors together from information that was collected from a review of the infant’s medical record (e.g., birthweight, gestational age, APGAR scores; for more information see [31]).

Maternal Depression

Depression was assessed using the Center for Epidemiologic Studies—Depression Scale (CES-D; [32]). The CES-D is a 20-item, self-report measure assessing for affective and somatic symptoms of depression during the past week. Each item is rated on a 4-point scale, with higher scores (range = 0 to 60) reflecting greater symptoms of depression. The CES-D demonstrated good to excellent internal consistency (α = 0.85 to 0.92) which is consistent with previous estimates for the general U.S. and Latinx populations [33, 34].

For the purpose of the current study, the timing of maternal depression was assessed by examining depressive symptoms (range = 0 to 60) separately during pregnancy and 6 months postpartum to determine their individual impact on child development outcomes. Severity of maternal depression was defined by using a CES-D cut-off score of 16 (low severity group < 16, high severity group ≥ 16), which is traditionally used for the initial screening of significant depressive symptoms [29]. The chronicity of maternal depression was defined by assessing how many times women met the criteria for high depression severity (CES-D ≥ 16) at each time point (range = 0 to 2 times). Per the recommended guidelines by Brennan and colleagues [18], we created three mutually exclusive groups using CES-D scores to counteract the potential issue of collinearity between depression severity and chronicity in our analyses. These groups represented measures of the severity (low vs. high) and chronicity (one report of at least high severity vs. two reports) of maternal depressive symptoms and were defined by the number of times that a participant scored 16 or higher on the CES-D. These criteria resulted in three groups: (a) neither severe nor chronic (CES-D score < 16 at both time points); (b) severe but not chronic (CES-D score ≥ 16 at one time point); and (c) both severe and chronic (CES-D score ≥ 16 at both time points).

Child Cognitive Development

The child’s cognitive development was assessed using the Mullen Scales of Early Learning: AGS Edition (MSEL; [35]). The MSEL is an individually administered comprehensive measure of cognitive functioning for infants and preschool aged children, from birth through 68 months of age. It assesses a child’s abilities in visual, linguistic, and motor domains, and distinguishes between receptive and expressive processing. The MSEL consists of four subscales (receptive language, expressive language, visual reception, and fine motor functioning). Raw scores from each subscale are standardized (T score) based on the child’s gender and age and has a population mean of 50 and a standard deviation of 10 [35]. Similar norms have been found for Spanish-speaking children up to 26 months of age [36]. A summary measure (i.e., Early Learning Composite Score) is derived from the standardized subscale scores, with a mean of 100 and a SD of 15, that represents the overall cognitive functioning of each child. This composite score is used to obtain the child’s percentile rank which represents what percent of children from their norm group (based on gender and age) that the child scored higher than on the MSEL (M_{percentile rank} = 50). The MSEL is a structured, untimed assessment that relies largely on non-verbal commands to assess cognitive levels. Previous studies have shown high test–retest and interrater reliability and high construct and concurrent validity of the MSEL for both English- and Spanish-speaking children [36], with internal consistency of the five scales for the current study ranging from 0.71 to 0.80 (α = 0.78). Higher scores indicate a higher level of cognitive development.

Child Social‑Emotional Functioning

Social-emotional functioning was assessed using the Vineland Social-Emotional Early Childhood Scales (VSEEC; [37]). The VSEEC is a semi-structured interview with the mother to assess the social-emotional development of children from birth through 71 months. It examines the child’s world of feelings and relationships and documents how he or she interacts in the home and external environment. The VSEEC consists of three subscales (play and leisure time, interpersonal relationships, and coping skills). These subscales can be summed to make a total composite score (i.e., Sum of Scale Standard Score), which represents the child’s overall social-emotional functioning. The subscales and total composite scores are standardized based on the child’s gender and age with a mean of 100 and a SD of 10 [37]. For children under 2 years of age, the coping skills subscale is not assessed nor included in the total composite score. The VSEEC is highly reliable and valid and has high test–retest reliability [37], with internal consistency of the four scales for the current study ranging from 0.71 to 0.87 (α = 0.80). Higher scores indicate higher levels of social-emotional development.

Statistical Analyses

An a priori power analysis was conducted using G*Power to ensure an adequate sample size was obtained [38]. Based on the medium effect size found in past research that examined the relationship between maternal depression and child development in low-income populations [21, 23], a sample size of approximately 91 participants was needed in order to obtain statistical power at the recommended 0.80 level at an alpha level of 0.05. This power analysis was conducted to detect both main (i.e., maternal depression) and interaction (i.e., child gender by maternal depression) effects on study outcomes.

Pearson product-moment correlations and independent samples t test analyses were used to identify possible covariates (i.e., gestational age, child gender, number of children, family income, maternal age, marital status, number of birth complications, and maternal and child language) on study outcomes. Only child gender (girl vs. boy) was significantly associated with child development outcomes (p < 0.05) and was included as a covariate in subsequent analyses.

Given the collinearity between maternal depression during pregnancy and 6 months postpartum (r = 0.37, p < 0.001) and these depression scores and severity/chronicity of maternal depression (r = 0.64 to 0.71, p < 0.001), these characteristics of maternal depression were tested in separate regression models. Specifically, hierarchical regression models were tested to identify whether the timing (pregnancy and 6 months postpartum CES-D scores) and the severity/chronicity of maternal depression (dummy coded with three levels: neither severe nor chronic, severe but not chronic, or both severe and chronic) were associated with child development outcomes (cognitive and social-emotional development), controlling for child gender. For each regression model, child gender was entered first followed by one of the depression variables (either the pregnancy depression score, 6 months postpartum depression score, or the severity/chronicity of maternal depression variables). A secondary regression model was also conducted to examine the unique contributions of the timing and severity/chronicity of maternal depression, controlling for child gender. The effect sizes for these associations were presented as Cohen’s f² with values of 0.02, 0.15, and 0.35 representing a small, medium, and large effect size, respectively [39]. Hierarchical regression analyses were also used to test for two-way interactions (i.e., child gender by maternal depression). Given the number of comparisons, a Bonferroni correction was conducted with interaction terms of p < 0.005 considered significant. Case-wise deletion of missing data was used in all analyses.

Results

Participants

Participants were approximately 24 years of age (SD = 4.49; range = 18–35 years), in their 17th week of pregnancy when they entered the study (SD = 5.09; range = 6–29 weeks), and had less than a high school education (M = 10th grade education, SD = 3.25). Most women were married or living with a partner (73%) and had an annual household income under $20,000 (92%) in 1998–2001. Of these participants, the majority were Latina women (88%) who were born in Mexico/Central America (77%), and were on average 16 years of age when they immigrated to the U.S. Non-Latina women were from various ethnic backgrounds (i.e., 7% African-American, 4% non-Hispanic White, and 1% Asian-American). Children born during the study were mostly boys (57%), had a Spanish-speaking mother (76%), were breastfed (90%), and on average, had experienced at least one medical complication at birth (49%; range = 0–6 birth complications; e.g. 43% needing oxygen at birth, 17% C-section delivery). The average age of the children when they participated in the developmental assessments was 32 months (range = 14–61 months). The majority of children had at least one sibling (58%).

Maternal Depression Characteristics—Pregnancy to 6 Months Postpartum

Mothers’ depressive symptoms significantly decreased from pregnancy (M_CES-D = 19.4, SD = 10.1) to 6 months postpartum [M_CES-D = 14.0, SD = 11.0; t(86) = 4.5, p < 0.001] with 66% of mothers having high depressive symptoms (CES-D score ≥ 16) during pregnancy compared to 39% at 6 months postpartum. In terms of depression severity and chronicity, 28% of mothers (n = 27) had depressive symptoms that were considered neither severe nor chronic (CES-D score < 16 at both time points). Another 43% of mothers (n = 41) had depressive symptoms that were considered severe but not chronic (CES-D score ≥ 16 at one time point; 80% occurred during pregnancy). Finally, 29% of mothers (n = 28) had depressive symptoms that were considered to be both severe and chronic (CES-D score ≥ 16 at both time points).

Impact of Maternal Depression on Child Cognitive Development

Approximately two-thirds of cognitive assessments (MSEL) were conducted in Spanish (63%), 20% in English, and 17% in both English and Spanish. On average, children had lower cognitive development scores (M = 86.3, SD = 14.4; range = 49–130) than the national norms (M = 100, SD = 15; [35]). They also had an average percentile rank score of 25 (SD = 25.0), meaning that they scored higher than 25% of other children from their norm group who took the MSEL (based on gender and age), which is lower than the national norms (M = 50; [35]). Compared to equivalent-aged children, 45% of children from the current sample fell within the “very low/below average” category of overall cognitive development, 50% were in the “average” category, and 5% were in the “above average/very high” category [35]. Lower cognitive development seemed to be mostly driven by lower average scores on receptive language (M = 36.6, SD = 8.3) and expressive language (M = 38.6, SD = 9.9), as opposed to visual reception (M = 45.4, SD = 10.3) and fine motor skills (M = 49.8, SD = 11.9), when compared to national norms on these subscales (M = 50, SD = 10; [35]).

Regression analyses revealed that child gender and maternal depressive symptoms experienced during pregnancy were significantly associated with child cognitive development (see Table 1). Child gender accounted for the greatest increment of variance for cognitive development, with 10% explained by its entry into the model; girls had higher cognitive development scores than boys (p = 0.004). Mothers with higher depressive symptoms during pregnancy had children with lower cognitive development scores at 1½ to 4½ years postpartum (p = 0.040), accounting for 5% of the variance. The overall variance explained by the final model was 15% (f² = 0.17). These effects remained even after including maternal depressive symptoms at 6 months postpartum and depression severity/chronicity in the model (R²Δ = 0.03, β = − 0.18, p = 0.030, f² = 0.17). Neither maternal depressive symptoms at 6 months postpartum (R²Δ = 0.01, β = − 0.09, p = 0.428, f² = 0.11) nor depression severity/chronicity (R²Δ = 0.03, β = − 0.16, p = 0.129, f² = 0.12) were significantly associated with child cognitive development (see Table 2 for child cognitive development scores by maternal depression severity/chronicity).

Table 1.

Hierarchical regression analyses for characteristics of maternal depression associated with child cognitive and social-emotional development at 1½ to 4½ years of age (N = 96)

Variables	Cognitive development			Social-emotional development
	R 2	Standard β	Standard error	R 2	Standard β	Standard error
	0.15**			0.07*
Child gender		− 0.30**	5.13		− 0.03	11.00
Prenatal depression		− 0.22*	0.26		− 0.27*	0.54
	0.12**			0.09*
Child gender		− 0.32**	5.28		0.02	11.51
6 months postpartum depression		− 0.09	0.24		− 0.30**	0.52
	0.14**			0.13**
Child gender		− 0.31**	5.19		− 0.04	10.81
Depression severity/chronicity
Neither severe/chronic		− 0.19	6.76		− 0.39**	14.07
Severe but not chronic		0.02	6.12		− 0.09	12.74
Both severe & chronic		0.07	5.66		0.08	12.74

^*p < 0.05,

^**p < 0.01

Table 2.

Child cognitive and social-emotional development scores by severity/chronicity of maternal depression

	Severity/chronicity of maternal depression
	Neither severe nor chronic n = 27	Severe but not chronic n = 41	Both severe and chronic n = 28
Cognitive development
Composite Standard Score M (SD)	88.59 (± 16.77)	84.87 (± 12.59)	86.27 (± 14.67)
Composite Percentile Rank Score M (SD)	33.47 (± 32.42)	21.96 (± 18.43)	22.79 (± 24.30)
Receptive Language T Score M (SD)	35.98 (± 8.72)	34.64 (± 7.89)c	39.51 (± 8.51)b
Expressive Language T Score M (SD)	39.11 (± 10.41)	37.18 (± 9.08)	40.13 (± 10.62)
Visual Reception T Score M (SD)	46.23 (± 9.98)	46.62 (± 10.24)	44.50 (± 10.31)
Fine Motor Skills T Score M (SD)	54.18 (± 14.21)c	49.55 (± 11.16)	47.21 (± 9.59)a
Social-emotional development
Sum of Scale Standard Score M (SD)	274.22 (± 46.45)bc	236.22 (± 52.36)a	226.93 (± 54.21)a
Composite Standard Score M (SD)	99.86 (± 14.18)	96.12 (± 15.62)	99.67 (± 17.50)
Play & Leisure Time T Score M (SD)	86.99 (± 14.03)	88.71 (± 16.72)	93.51 (± 15.03)
Interpersonal Rel. T Score M (SD)	100.23 (± 15.29)	97.89 (± 18.43)	100.83 (± 19.77)
Coping Skills T Score M (SD)	111.54 (± 10.12)	110.84 (± 11.16)	113.02 (± 11.27)

ANCOVAs were conducted to examine for between group differences (severity/chronicity of maternal depression groups) on cognitive and social-emotional development outcomes, controlling for child gender

^aCompared with Neither Severe nor Chronic group (CES-D score < 16 at both time points), p < .01

^bCompared with Severe but not Chronic group (CES-D score ≥ 16 at one time point), p < .01

^cCompared with Both Severe and Chronic group (CES-D score ≥ 16 at both time points), p < .01

A significant child gender by prenatal depression interaction was also found such that girls born to mothers with higher prenatal depressive symptoms had lower cognitive development scores at 1½ to 4½ years postpartum compared to girls born to mothers with lower prenatal depressive symptoms (R² = 0.19, β = −0.22, p = 0.001, f² = 0.24). Boys had lower levels of cognitive development regardless of their mother’s prenatal depression levels (see Fig. 2).

Fig. 2.

Interaction of maternal depression during pregnancy and child gender on child cognitive development at 1½ to 4½ years of age

Child Social‑Emotional Development

On average, children had similar social-emotional development scores (M = 98.2, SD = 15.8; range = 59–143) as the national norms (M = 100, SD = 10; [37]). Compared to equivalent-aged children, 22% of children from the current sample fell within the “moderately low or low” category of overall social-emotional development, 68% were in the “adequate” category, and 10% were in the “moderately high or high” category [37]. Lower social-emotional development seemed to be mostly driven by lower average scores on play and leisure time (M = 89.7, SD = 15.6) versus interpersonal relationships (M = 99.4, SD = 17.9) and coping skills (M = 111.5, SD = 10.5), when compared to national norms on these subscales (M = 100, SD = 10; [37]).

Maternal depressive symptoms experienced during pregnancy and at 6 months postpartum, as well as depression severity/chronicity were significantly associated with child social-emotional development at 1½ to 4½ years postpartum (see Table 1). Mothers with higher depressive symptoms during pregnancy (R² = 0.07, β = − 0.27, p = 0.010, f² = 0.08) and at 6 months postpartum (R² = 0.09, β = − 0.30, p = 0.005, f² = 0.10) had children with lower social-emotional development scores, accounting for 7% and 9% of the variance, respectively. Mothers with depressive symptoms that were considered to be more severe or chronic (CES-D score ≥ 16 at one or both time points) had children with lower social-emotional development scores (R² = 0.13, β = − 0.39, p = 0.001, f² = 0.12), accounting for 12% of the variance (see Table 2 for child social-emotional development scores by maternal depression severity/chronicity). A secondary regression model (R² = 0.12, f² = 0.14) revealed that only maternal depressive symptoms at 6 months postpartum remained significantly associated with lower child social-emotional development scores (R²Δ = 0.05, β = − 0.31, p = 0.031) when compared to prenatal depression (R²Δ = 0.06, β = − 0.17, p = 0.143) and depression severity/chronicity (R²Δ = 0.01, β = − 0.11, p = 0.459). Child gender was not significantly associated with child social-emotional development. No significant two-way interaction was found for child gender by maternal depression on child social-emotional development (R² = 0.12, β = − 0.07, p = 0.552).

Discussion

Although a significant body of literature has shown that maternal depression can adversely affect child development, few studies have prospectively examined how different characteristics of maternal depression may be associated with developmental outcomes among low-income families. The current prospective study is one of the first to examine whether the timing (pregnancy vs. early postpartum), severity, and chronicity of maternal depression were associated with child cognitive and social-emotional developmental outcomes in low-income mothers and their children (the majority of mothers who were Latinx immigrants to the U.S.).

Regarding child cognitive development, results indicated that women who experienced higher depressive symptoms during pregnancy (compared to depression at 6 months postpartum and depression severity/chronicity) had children with lower levels of cognitive development at 1½ to 4½ years postpartum. Few prospective studies have examined the effects of prenatal depression on cognitive development with seemingly contradictory results of no effects found to medium effect sizes [8, 11]. These inconsistencies may reflect, among other things, methodological differences in when depression is assessed (e.g., early vs. late pregnancy) and how child development is defined (i.e., use of standardized measures vs. parent self-report). Results from the current study support the adverse impact that high maternal depressive symptoms experienced during early pregnancy (first and second trimester) may have on the cognitive development of low-income children. Researchers have theorized that experiencing depressive symptoms during pregnancy may indirectly affect cognitive and language development during childhood by affecting fetal development. More specifically, prenatal depression may program the developing fetal hypothalamic pituitary adrenal (HPA) axis to be more reactive to stress, resulting in altered patterns of the stress hormone cortisol later in childhood, causing children to become more easily over-aroused in different learning situations and interfering with their development of executive functioning and memory consolidation [9]. Indeed, a prospective study by Davis & Sandman [40] demonstrated that elevated maternal cortisol levels during the first trimester of pregnancy, compared to the third trimester, was predictive of lower child cognitive development by 12 months of age. Given the unique stressors often experienced by low-income and Latina immigrant mothers (e.g., acculturative stress, discrimination, limited access to health care services), these results support the need for depression prevention interventions during the early stages of pregnancy that are designed to minimize these potentially long-term adverse effects on child cognitive development.

Results from the current study also revealed that higher maternal depressive symptoms during pregnancy especially affected the cognitive development of girls compared to boys, with boys having lower levels of cognitive development regardless of their mother’s prenatal depression levels. These results are consistent with that of previous studies showing that boys generally display lower levels of cognitive development compared to girls and may reflect how prenatal exposure to elevated levels of maternal stress biomarkers, such as cortisol, differentially influences the neurodevelopment of boys versus girls [9]. For instance, recent pregnancy studies have shown that male fetuses exposed to elevated levels of maternal cortisol early in pregnancy exhibit delayed physical and neuromuscular development that impact their general cognitive ability after birth [40, 41]. Among female fetuses, studies show that exposure to elevated levels of maternal cortisol during early pregnancy result in larger right amygdala volume, which has been associated with greater sensitivity to maternal depression after birth, resulting in long-term effects on their emotion regulation and learning [20, 42]. These gender differences in fetal programming may be exacerbated by contextual factors (e.g., socioeconomic status) that impact the cognitive development of children from low-income and Latinx immigrant families, as socioeconomic status may affect the amount of cognitive stimulation that children receive at home depending on the education and resources available to the parents [21, 22]. However, these cognitive deficits have been shown to be reversible among low-income children of parents who participate in Head Start programs that focus on teaching linguistic and cognitive stimulation activities to practice at home with their children [43]. Such programs have shown demonstrable improvements in language-based measures of cognitive development among children of low-income Latinx families [43]. Yet, further studies are needed to evaluate the impact of such cognitive stimulation programs on the mechanisms that may predispose low-income children to cognitive delays later in childhood.

Results also indicated that both the timing and severity/chronicity of maternal depression were associated with child social-emotional development. More specifically, women who experienced higher depressive symptoms during pregnancy and at 6 months postpartum, and those with higher depression severity/chronicity had children with lower levels of social-emotional development at 1½ to 4½ years postpartum. These results are consistent with recent prospective studies in Norway and Australia that found higher maternal depressive symptoms during pregnancy, during early postpartum, and at both time points to be associated with lower social-emotional development of children up to 2 and 4 years of age [14, 15]. Results of these studies suggest that both pre- and postnatal depression can independently impact children’s social-emotional development through different pathways. As previously mentioned, prenatal depression may affect children’s social-emotional development by programming the developing fetus’ HPA axis to be more reactive to stress, resulting in children becoming more easily over-aroused across a variety of non-threatening situations after birth, including interacting and playing with others, that may also impact their ability to cope in novel situations [9]. At 6 months postpartum, infants begin to learn the routines and predictable responses of human social interaction through their mothers. Therefore, maternal depression during this critical period of development may impact children’s social-emotional development through inconsistent maternal responses to infant cues that result in missed opportunities for mothers to promote their child’s learning of social responses and behaviors, as well as insecure infant-mother attachment [9]. In turn, severe and chronic maternal depressive symptoms across both time points can have long-lasting detrimental effects on children’s social-emotional development during the preschool years due to mother–child interactions that do not adequately model emotion regulation and coping skills, characterized by greater negativity, flat affect, social withdrawal, and poor communication [9, 44]. Such effects have been shown to be particularly salient for the social-emotional development of low-income, ethnic minority children [7, 23], highlighting the importance of screening for and preventing and treating depressive symptoms during pregnancy and the postpartum period in this population. The U.S. Preventive Services Task Force specifically recommends that clinicians provide or refer pregnant or postpartum women who are at increased risk of perinatal depression to preventive interventions [45].

Study Limitations

The current results should be interpreted with some degree of caution given several study limitations. First, given that the majority of our sample consisted of low-income, Latina immigrant mothers (77%) and their children (63% of whom primarily spoke Spanish), the results may not be generalizable to mothers and children from other socioeconomic or ethnic backgrounds. For instance, it should be noted that lower cognitive development in the current study seemed to be mostly driven by language deficits (i.e., lower scores on receptive and expressive language), as opposed to broader cognitive delays. This pattern of results may be a reflection of the lack of available Spanish-language norms for older children (> 26 months of age) of Latinx families on the Mullen Scales of Early Learning and highlight the need for empirical studies to establish such norms. Nevertheless, this is one of the first studies to examine the long-term impact of maternal depression on child development in this growing and important at-risk population. Second, the sample size was relatively small (n = 96), as this was a subsample from two separate cohorts of pregnant women (recruited from 1998 to 2001) who had participated in studies focused on identifying and preventing the onset of major depression among low-income, pregnant women [29, 30]. As such, the sample consisted of children ranging from 14 to 61 months of age. However, this study was adequately powered and our results showed medium effect sizes (f² = 0.12 to 0.24) when examining the association of the timing and severity/chronicity of maternal depression on standardized scores of child cognitive and social-emotional development, which is consistent with the effect sizes found in previous studies with low-income families [21, 23]. The strength of our findings across this wide child age range may be due to the precision and stability of the standardized development scores used improving as children get older and as they are further removed from the effects of prenatal depression, but not necessarily the continued impact of maternal depression during early childhood. Additional studies of more recent cohorts in this research area, with larger sample sizes, would aid in supporting our findings. Such studies would also further our understanding of potential mediators (e.g., prenatal cortisol, mother–child interactions, birth complications) and moderators (e.g., ethnicity, acculturation) that may influence maternal depression effects on child development among low-income immigrant mothers and their children over time. Third, the assessment of prenatal depression, while significantly associated with child development outcomes in the current study, was conducted sometime between the 6th and 29th weeks of pregnancy, with most mothers assessed during the first and second trimesters. Multiple assessments of depressive symptoms in each trimester of pregnancy are indicated to more thoroughly assess the effects of timing of maternal depression during pregnancy on child development. Finally, although a highly reliable and valid standardized measure of social-emotional development was used (Vineland Social-Emotional Early Childhood Scales), it was based on mothers’ self-report. Therefore, mothers experiencing depressive symptoms may have been more negatively biased when rating their children. Future studies should include observational assessments of children’s social interactions with same-aged peers and their coping skills during challenging situations.

Summary

Our results suggest that maternal depression experienced during pregnancy has long-term adverse effects on child cognitive development up to 5 years postpartum, particularly among girls. Additionally, both the timing (pregnancy and early postpartum) and severity/chronicity of maternal depression are each independently associated with child social-emotional development in this low-income sample. These results highlight the importance of examining the impact of contextual risk factors (e.g., limited socioeconomic resources) that may interact with biological mechanisms (e.g., HPA axis regulation) during pregnancy to increase mothers’ risk of developing depression and subsequently impacting fetal programming of their infants’ brain development and stress response system in utero. These shifts during pregnancy, together with exposure to maternal depressive behaviors during the early postpartum period (e.g., poor mother-infant interactions), may affect infants’ cognitive and social-emotional processing that continues later in childhood. Given the adverse effects that maternal depression has been shown to have on child development, particularly among low-income mothers and their children, pre- and postnatal preventive and treatment interventions are needed to help women prepare for the many challenges related to pregnancy, childbirth, and motherhood. Such interventions should be tailored to the cultural resources and strengths that women and their families have to manage pre- and postnatal depressive symptoms, as well as to promote learning and emotion regulation in their children [46]. Current evidence shows that depression can be prevented [47, 48]. Randomized controlled studies are now needed to demonstrate the potential long-term benefits that these depression prevention interventions may have for mothers and their children, as well as to examine whether these benefits are a result of improvements in cortisol regulation and the quality of the social interactions that a mother has with her child. Preventing perinatal depression should be one of the highest public health priorities because it benefits both the mother and the developing child [49].