Prevalence and Risk Factors of Maternal Depression During the First Three Years of Child Rearing

Abstract

Background:

This longitudinal study examined maternal depression status from birth of a child to 36 months of age using data from the National Institute of Child Health and Human Development (NICHD) Study of Early Child Care and Youth Development.

Methods:

Maternal depression was assessed using the Center of Epidemiological Studies Depression Scale (CES-D) and defined as a score of ≥16. For this study, early onset depression was defined as depression within the 6 months after birth, and late onset depression was depression onset when the child was ≥24 months old. Chronic depression was defined as depression that started within 6 months after birth and lasted until 24 months of age or longer.

Results:

The prevalence of maternal depression was 32.2% for early onset, 7.4% for late onset, and 13.4% for chronic depression. The prevalence of maternal depression was highest at 1 month, decreased at 6 months, and then remained fairly stable until 36 months. Mothers 18–24 years of age, of black race, unemployed, with lower social support, single, or with poor general health had a higher prevalence of both early and chronic depression compared to other groups.

Conclusions:

Younger maternal age, poverty, lower education, and lack of social support were all significantly associated with increased maternal depression in multivariate regression models. Younger age, black race, unemployment, single status, lack of social support, and poor general health were all risk factors for increased prevalence of maternal depression.

Introduction

Depression is one of the top five disabling disorders worldwide¹ and the leading cause of disease-related disability among women.² Major depression has been the leading cause of years lived with disability and the fourth leading cause of burden among all diseases. For women aged 15–44 years, depression is the second leading cause of disability, exceeded only by HIV/AIDS.³ In the United States, depression is the leading cause of nonobstetrical hospitalization among women aged 18–44 years. In 2000, >200,000 women aged 18–44 years were discharged with a diagnosis of depression; 7% of all hospitalizations among young women were for depression.⁴ Although depression affects both men and women, women are more likely to be diagnosed with depression in any given year.⁵ Women have an estimated lifetime risk of 14% for depression, with a slightly higher risk of 14.5% in the first 3 months after childbirth.⁶ A meta-analysis of studies from developed countries showed that almost one fifth of women had depression during the first 3 months postpartum.⁷ For young and socioeconomically disadvantaged mothers, the risk is even greater, at about 25%.⁸ Compared to mothers without depression, depressed mothers have a 3-fold greater risk of their children having serious emotional problems, and there is a 10-fold greater risk of depressed mothers having a poor relationship with their children.⁹

Maternal depression has been studied as a risk factor for impaired caretaking capacity; included are insensitive, nonresponsive, and nonstimulating care that can lead to adverse outcomes of infants' and children's psychological development, intellectual competence, psychosocial functioning, rate of psychiatric morbidity, and physical well-being.^10,11 Interestingly, Weissman et al.¹² have suggested that remission of maternal depression may be associated with reductions in the mental and behavioral disorders of their children. Although evidence supports that maternal depression is a significant public health problem, there are few long-term follow-up studies of mothers after they give birth. Therefore, studies that evaluate the risk factors for maternal depression over the different periods of child rearing are needed. The main purposes of this study are to (1) examine the prevalence of maternal depression status for different periods from the birth of the child to 36 months of age, (2) evaluate several characteristics as potential risk factors for maternal depression, and (3) determine the rates of early onset, late onset, and chronic depression.

Materials and Methods

Data resource and study sample

Data for this study came from the National Institute of Child Health and Human Development (NICHD) Study of Early Child Care and Youth Development (SECCYD). Participants in the NICHD SECCYD were recruited from hospitals at 10 research sites from 10 different states. Recruitment began in January 1991 and was completed in November 1991. NICHD SECCYD was one of the most comprehensive longitudinal studies of children initiated by the NICHD. A total of 1364 families were enrolled, and participants were selected by conditional random sampling. The sampling plan was designed to ensure the following distributions: (1) 60% were mothers who planned to work or to go to school full-time during the child's first year, (2) 20% were mothers who planned to work or attend school part-time during the child's first year, and (3) 20% were mothers who planned to stay at home full-time with the child. The selected families demonstrated the demographic diversity of the sites from the economic, educational, and ethnic perspectives and included both two-parent and single-parent families. Major exclusions were (1) mothers <18 years of age at the time of the child's birth, (2) mothers who did not anticipate remaining in the catchment area for at least 3 years, (3) children with obvious disabilities at birth or who remained in the hospital more than 7 days postpartum, and (4) mothers not sufficiently conversant in English.

Beginning with the time of enrollment (the 1-month home visit), families were scheduled for multiple periodic visits for data collection. Our study used data from phase I of the NICHD SECCYD study, which was through the first 36 months of the child's life. During phase I, there were site visits to each child at home, in child care (if used), and in a laboratory playroom. Other phases of NICHD SECCYD were not included in our study and were phase II (54 months through first grade), phase III (2nd through 6th grades), and phase IV (ages 14 and 15).

Variables of interest

Maternal age and ethnicity

Maternal age and ethnicity were obtained at the time of the site visit when the child was 1 month old. Maternal age was categorized into the following intervals: 18–24, 25–34, and 35–46 years old. Initially, five categories of race and ethnicity were used: American Indian, Eskimo, or Aleut; Asian or Pacific Islander; black or African American; white; and other. Because of small numbers in some categories, only three categories were used for analysis: whites, blacks, and other (all those who are neither black nor white).

Family income and poverty

Family income level was measured as the ratio of income to needs, calculated as the total family income divided by the poverty threshold for their family size (the federal poverty line for a family of four in 1991 was $13,924). If ratio of income to needs was <1, the household was considered poor, and if the ratio was ≥1, the household was considered not poor.

Maternal marital status

Marital status was determined from the mother at the visit when the child was 1 month old and was categorized as (1) mother single, not married, (2) mother separated or divorced; and (3) mother currently married. Marital status was collapsed into two categories, single and not single, because of the small numbers of subjects in some of the original categories.

Maternal employment and maternal education

These were obtained at the time of the 1 month of age visit. Maternal employment status was categorized into the following groups: not employed, employed and at work, employed and on leave. Maternal education was divided into two categories: less than bachelor's degree and bachelor's degree or above. Two categories were used because of small numbers in some categories. Dividing into two categories allowed use of the t test for analysis.

Maternal general health

To assess maternal general health, mothers were asked to describe their health compared to that of other women of their age. Respondents answered on a 4-point scale: 1=poor health, 2=fair health, 3=good health, 4=excellent health. This question was asked at multiple visits; the information obtained when the child was 1 month old was used as the baseline for this study. Based on the small numbers in the poor health category, maternal health was divided into three categories: 1=poor or fair health, 2=good health, 3=excellent health.

Social support

The Relationships with Other People questionnaire was used to measure both general and perceived availability of social support and was based on how respondents rated their relationships over the past month, considered the most consistent predictor of adult support.¹³ This self-administered questionnaire consisted of 11 items rated on a 6-point scale with 1=none of the time and 6=all of the time. A composite score was formed for each subject as the imputed mean of the 11 questionnaire items. Scores could range from 1 to 6, with higher values indicating more social support. Cronbach's alpha indicates high levels of internal consistency within the set of items, which were based on Weiss's concepts of social relationships.¹⁴ For comparison, three categories were defined: low social support (<4.82), moderate social support (4.82–5.64), and high social support (>5.64).

Maternal depression

Maternal depression was assessed using the Center for Epidemiological Studies Depression scale (CES-D) and was determined at the 1, 6, 15, 24, and 36 months home visits. The CES-D is a 20-item, self-report depression scale developed to identify depression in the general population.¹⁵ The scale is one of the best-known survey instruments for identifying symptoms of depression, has norms available, and has been used extensively in large studies.¹⁶ In line with the work of Radloff's group,¹⁷ the CES-D shows excellent internal consistency (Cronbach's alpha >0.85) and test-retest correlation >0.5. A cutoff score of ≥16 was used as indicative of depression in this study. Depression status was further defined as early onset (occurring within 6 months after giving birth), late onset (occurring when the child was at least 24 months old), and chronic depression (depression lasting from early onset to late onset or beyond).

Statistical analysis

The prevalence of maternal depression was determined by gender of child, maternal age group, maternal employment status, and race. Descriptive statistics were used to describe maternal depression at various periods postpartum. Bivariate analyses were done using such tests as the t test and chi-square test as appropriate to compare maternal depression at different times. Initial bivariate analyses of the independent variables (including potential risk factors) and outcome variables were used to determine if there was a significant association between each independent variable and each outcome variable. The independent variables included maternal age, race, family income/poverty, maternal marital status, maternal employment, maternal education, maternal general health, and social support. The dependent (outcome) variable was maternal depression.

Multiple logistic regression models were used to assess the associations of multiple potential risk factors simultaneously for each of the binary outcome variables of early onset, late onset, and chronic depression. Each outcome was evaluated in a separate model. As maternal depression status was assessed repeatedly at the child's age of 1, 6, 15, 24, and 36 months, the generalized estimating equations (GEE), which provides support for such nonindependent data as repeated measures and clustered data, was used for analyses. The GEE model is an extension of the logistic regression model for correlated responses and takes into account that maternal depression is repeatedly examined in the course of the study. This model was used to assess the relationship of potential risk factors with the maternal depression status repeatedly measured over time. SPSS version 16 was used for all data analyses (SPSS Chicago, IL).

Results

Characteristics of study participants

Maternal characteristics at the visit when the child was 1 month old are shown in Table 1. Mean maternal age was 28.1 years (standard deviation [SD] 5.63 years), with most (86.7%) women <34 years old; 82.6% were white. Only 14.4% were single (i.e., not married or living with their partner), and 22.7% were considered poor.

Table 1.

Characteristics of Participants: Information Attained 1 Month After Birth

	Participants
Characteristic	n	Proportion (%)
Maternal age
18–24	386	28.3
25–34	796	58.4
35–46	182	13.3
Maternal ethnicity
White	1,127	82.6
Black	174	12.8
Other	63	4.6
Marital status
Not single	1,166	85.5
Single	196	14.4
Poverty status
Poor	964	22.7
Not poor	310	70.7
Child gender
Male	705	51.7
Female	659	48.3
Mother's employment status
Employed in last 12 months	536	39.3
Not employed in last 12 months but now employed	135	9.9
Not employed in last 12 months and not now employed	693	50.8
Mother's education
Less than bachelor's degree	881	64.6
Bachelor's degree or above	482	35.3

Prevalence of maternal depression

Overall, the prevalence of maternal depression was 25.6% at 1 month, 16.3% at 6 months, 15.4% at 15 months, 15.0% at 24 months, and 15.0% at 36 months of the child's age. Prevalence of depression for subgroups is shown in Table 2. The prevalence of depression was consistently highest for mothers <24 years old compared to the other maternal age groups at all assessment time points from the 1 to 36 months of age visits. Women with the maternal characteristics of black race and poor had a higher prevalence of depression. The prevalence for depression was higher for mothers who were single, unemployed, had less than a bachelor's degree, and had low social support for all periods of the child's age. The prevalence of depression was lower when mothers reported a better health status.

Table 2.

Prevalence of Maternal Depression at Different Months of Child Rearing by Variables of Interest

Child's age (months) at time of assessment
	1			6			15			24			36
	n	%	p	n	%	p	n	%	p	n	%	p	n	%	p
Total	349	25.6		222	16.3		210	15.4		205	15.0		205	15.0
Maternal age, years			^**			^**			^**			^**			^**
18–24	144	37.4		93	27.6		92	28.4		88	33.0		80	25.9
25–34	185	20.7		116	13.5		104	12.4		107	13.8		118	14.5
35–46	20	23.3		13	15.5		14	17.3		10	13.5		7	8.8
Ethnicity			^**			^**			^**			^**
Black	68	39.1		46	29.7		39	26.5		37	29.8		31	22.8
White	266	23.6		164	15.4		163	15.7		157	16.6		165	16.3
Other	15	23.8		12	21.1		8	14.0		11	23.4		9	16.4
Poverty status			^**			^**			^**			^**			^**
Not poor	193	20.0		114	12.3		121	13.3		117	14.0		119	13.5
Poor	126	40.8		81	30.1		73	28.7		67	30.3		66	26.7
Marital status			^**			^**			^**			^**			^**
Not single	268	23.8		163	14.7		163	15.1		154	15.7		158	15.1
Single	80	40.8		58	34.1		46	28.4		51	37.2		47	29.9
Employment status			^**			^**			^**						^**
Not employed	165	30.8		121	24.8		112	23.6		86	20.5		99	21.7
Employed and at work	19	14.1		17	13.1		15	12.2		19	16.7		14	11.4
Employed and on leave	165	23.8		84	12.7		83	12.9		100	17.1		92	14.8
Education			^**			^**			^**			^**			^**
Less than bachelor	264	30.0		171	21.1		166	21.2		164	24.0		155	20.6
Bachelor or above	85	17.6		51	10.9		44	9.6		41	9.4		50	11.1
Maternal general health			^**			^**			^**			^**			^**
Poor or fair	37	55.2		31	52.5		27	48.2		21	45.7		20	37.7
Good	183	31.0		118	21.4		107	20.1		103	21.5		113	21.7
Excellent	129	18.3		73	10.9		76	11.7		81	13.6		72	11.5
Social support			^**			^**			^**			^**			^**
Low	131	45.2		88	33.3		70	28.1		66	29.6		66	28.1
Median	161	22.7		98	14.4		101	15.2		98	16.3		103	15.8
High	57	15.7		36	10.7		39	11.9		41	13.9		36	11.4

Group comparison is significant at the 0.01 level (2-tailed).

Early onset, late onset, and chronic maternal depression

Table 3 shows the prevalence of early onset, late onset, and chronic maternal depression for different maternal characteristics of interest. The prevalence of early onset, late onset, and chronic maternal depression was 32.2%, 7.4%, and 13.4%, respectively. All the examined maternal characteristics were found to be significantly associated with both early onset and chronic depression but not with late onset depression. Mothers who were 18–24 years old, had a poor economic status, were of black race, were not employed, and reported lower social support and education were at a higher risk of having either early onset depression or chronic depression compared to mothers in other categories. Mothers who were single or who had poor general health had a higher risk for early onset depression or chronic depression in comparison to mothers who were not single or who had better general health.

Table 3.

Prevalence of Early Onset, Late Onset, and Chronic Maternal Depression by Variables of Interest

	Early	onset		Late	onset		Chronic	depression
	n	%	p	n	%	p	n	%	p
Total	439	32.2		101	7.4		180	13.4
Maternal age			^*						^*
18–24	178	49.7		22	6.8		91	28.9
25–34	211	27.4		65	9.1		78	10.5
35–46	50	27.8		14	8.4		14	8.0
Ethnicity			^*						^*
Black	83	50.9		13	8.9		34	23.0
White	334	30.8		86	8.5		138	13.4
Other	22	37.3		2	3.9		11	20.0
Poverty status			^*						^*
Not poor	247	26.3		71	8.2		94	10.3
Poor	151	53.2		20	7.7		71	28.4
Marital status			^*						^*
Not single	339	30.1		81	7.8		138	12.8
Single	99	55.3		20	12.0		45	29.4
Employment status			^*						^*
Not employed	213	42.3		30	6.4		92	19.8
Employed and at work	31	23.8		11	9.2		15	12.0
Employed and on leave	195	29.0		60	9.8		76	11.8
Education			^*						^*
Less than bachelor	326	39.0		71	9.3		149	19.3
Bachelor or above	113	24.0		30	6.8		34	7.4
Maternal general health			^*						^*
Poor or Fair	19	51.4		2	6.3		9	26.5
Good	134	40.2		21	7.0		57	18.3
Excellent	286	30.5		78	8.9		117	13.2
Social support			^*						^*
Low	155	56.0		20	7.8		68	27.9
Median	209	30.3		56	8.7		88	13.4
High	75	22.1		25	8.2		27	8.2

Significant at the 0.05 level (2-tailed).

Multiple variable analyses of the risk factors

Table 4 shows the multivariate analyses using logistic regression and GEE. After controlling for confounding with logistic regression, maternal age, employment status, and social support demonstrated negative or inverse associations with the early onset of maternal depression. Poverty showed a positive association with early onset maternal depression. Compared to mothers that were 18–24 years old, mothers 25–34 years old were at two times the risk for late onset depression. Maternal education status had a significant inverse association with late onset maternal depression; that is, better educated mothers were less likely to have late onset depression. Maternal age and education had a statistically significant inverse relationship to chronic maternal depression. Maternal report of higher social support was associated with a decreased risk of chronic postpartum depression. The risk factors of maternal depression revealed in GEE analysis were generally consistent with those for early and chronic depression found in the logistic models.

Table 4.

Risk Factors for Maternal Depression: Multiple Logistic Regression and Generalized Estimating Equation Models

	Multiple logistic regression						GEE
	Early onset		Late onset		Chronic depression		Maternal depression
	OR	95% CI	OR	95% CI	OR	95% CI	OR	95% CI
Maternal age
18–24	Ref		Ref		Ref		Ref
25–34	0.609^*	0.436-0.850	1.980^*	1.041-3.768	0.423^*	0.276-0.648	0.624^*	0.481-0.808
35–46	0.705	0.443-1.120	1.898	0.820-4.393	0.374^*	0.190-0.734	0.696^*	0.490-0.988
Ethnicity
White	Ref		Ref		Ref		Ref
Black	1.442	0.952-2.182	0.706	0.307-1.622	0.979	0.580-1.654	1.167	0.854-1.594
Other	0.864	0.451-1.654	0.534	0.125-2.275	1.135	0.505-2.552	0.866	0.512-1.467
Poverty status
Not poor	Ref		Ref		Ref		Ref
Poor	1.720^*	1.187-2.492	1.029	0.509-2.083	1.614^*	1.006-2.590	1.435^*	1.060-1.944
Marital status
Not single	Ref		Ref		Ref		Ref
Single	1.263	0.772-2.068	1.768	0.738-4.237	1.276	0.715-2.277	1.338	0.942-1.902
Employment status
Not employed	Ref		Ref		Ref		Ref
Employed and at work	0.623^*	0.388-1.000	1.577	0.744-3.346	0.858	0.455-1.619	0.690	0.469-1.016
Employed and on leave	0.806	0.610-1.064	1.583	0.959-2.615	0.915	0.625-1.340	0.864	0.696-1.072
Education
Less than bachelor	Ref		Ref		Ref		Ref
Bachelor or above	0.925	0.680-1.258	0.593^*	0.360-0.977	0.686	0.432-1.090	0.0733^*	0.575-0.934
Maternal general health
Excellent	Ref		Ref		Ref		Ref
Poor or fair	1.777	0.845-3.733	0.801	0.182-3.527	1.568	0.637-3.861	1.548	0.859-2.787
Good	1.291	0.966-1.726	0.730	0.423-1.257	1.049	0.706-1.558	1.061	0.846-1.330
Social support
High	Ref		Ref		Ref		Ref
Low	3.381^*	2.310-4.949	0.914	0.467-1.788	3.167^*	1.874-5.353	2.657^*	1.985-3.557
Median	1.525^*	1.103-2.109	1.062	0.632-1.784	1.759^*	1.093-2.833	1.433^*	1.112-1.845

Significant at the 0.05 level (2-tailed).

CI, confidence interval; OR, odds ratio; Ref, reference group; GEE, generalized estimating equations.

Discussion

This study examined the prevalence trends of maternal depression from birth of the child up to 36 months of age. The prevalence of maternal depression was highest at 1 month and then decreased until 6 months of age. From 6 months until the final time period examined at the child's age of 36 months, the prevalence of maternal depression remained fairly constant. Two possible explanations are considered: (1) Perhaps primipara mothers are more likely to become depressed during their first month after the birth, and it may take some time, up to 6 months, for them to accommodate to the new situation. Beyond that time they may have found ways to accept and deal with the new situation. (2) Perhaps multipara mothers may feel the stress of dealing with more than one child immediately after the birth of the child, but just as for primipara mothers, they develop ways to deal with and accept the situation over time. For both explanations, increasing the mothers' social support can help develop acceptance and confidence in their new role. Cox et al.¹⁸ found that psychiatric illness is more likely to occur during the first year after birth, and at least half of the cases occur in the first 3 months. Also, recent studies have provided supportive evidence that pregnancy and childbirth are critical periods that are associated with mothers being more susceptible to mental health problems.^19,20

A unique feature of this study was the examination of maternal depression at different times after the birth of a child. This enabled us to examine the prevalence trends for maternal depression during the first 3 years after birth. We found the prevalence of maternal depression during the first 3 years after birth was higher in younger as compared to older mothers. One possible explanation is that younger mothers are just not as prepared to deal with the stresses of being a mother and may also be more likely to have financial or time problems than women who are older and more established. Weillings et al.²¹ suggested that young mothers and their children face poorer prospects in life than do women who delay motherhood until they are older. Patterns of women having children at an early age tended to be repeated in subsequent generations. It has also been reported that women who develop depression before the age of 22 are less likely to graduate from college and are more likely to earn 12%–18% less than women who have the onset of major depressive disorder (MDD) after age 21 or do not develop depression at all.²²

Our study found that mothers who are black and poor were associated with an increased prevalence of depression compared to women in other categories. Onozawa et al.²³ found that the incidence of maternal depression after birth was higher in minority women, including black, Asian, and other minorities, compared to white women, with 27% and 15%, respectively. Peterson and Hawley²⁴ found that low income and worry over meeting financial obligations can be a significant stressor for mothers. The lack of adequate income was associated with poorer parental coping skills, lower family cohesion, less empathy, and inability to reverse roles within the family. On the other hand, higher family income provides the family with more options, including better access to healthcare, recreation, and other commodities. Although Barnet et al.²⁵ found that maternal age and socioeconomic status (SES) did not have a significant relationship with maternal depression symptoms, our study found significant relationships with both characteristics.

We found that single mothers had a higher prevalence of depression than nonsingle mothers during the first 3 years after birth. Prior studies have reported that single mothers have two to three times the prevalence of depressive symptoms as the general population.^26
–28Our study found that unemployed women had a higher prevalence of depression than employed mothers. This is consistent with other studies that show employed, as compared to nonemployed, mothers have financial independence, improved self-esteem, and increased access to social networks; all these factors enhance mothers' psychological well-being.^29

–32 Employment, regardless of income or number of hours worked, offers some protection against mental health problems.^29,31

We found that mothers without a college degree had a higher prevalence of depression than mothers with a college degree. It is a reasonable expectation that a higher education will provide more employment opportunities and fewer financial problems. Perhaps better maternal education also is associated with a better understanding of child development, and this may contribute to a lower prevalence of depression. For example, Judge³³ found a significant relationship between maternal education level and a family's efforts to be active and innovative and to seek new experiences.

In our study, mothers with higher social support had a lower prevalence of depression than mothers who received lower social support. Social support can be considered:

… an interaction process between people or groups of people, who through systematic contact establish bonds of friendship and information, receiving material, emotional, affective support, contributing to mutual well-being and constructing positive factors in prevention and health maintenance. Social support highlights the role individuals can play to solve daily situations in crisis moments.³⁴

Panzarine et al.³⁵ found that certain dimensions of social support were significantly associated with depressive symptoms. Hudson et al.³⁶ revealed a negative correlation between the mother's depressive scores and social support. A study by Reid and Meadows-Olive³⁷ found that mothers who report more feelings of loneliness were more likely to report depressive symptoms. Liabsuetrakul et al.³⁸ showed that anxiety and the lack of social support were the most important predictors of postpartum depression in Thai women. Other studies have supported the negative relationship seen between social support and maternal depression.^25,36,39 Using a longitudinal multivariate assessment to evaluate predictors of maternal depression, our study provides strong support for the negative relationship between social support and maternal depression. We believe that there is strong evidence that increased social support is associated with decreased risk of maternal depression.

After adjusting for confounders, maternal ethnicity, maternal general health, marital status, and employment status were found to not be significantly associated with maternal depression in our study. Barnet et al.,²⁵ Caldwell and Antonucci,⁴⁰ and Birkeland et al.⁴¹ pointed out that there were no significant differences in depressive symptoms when comparing the major ethnic groups. This is similar to our findings that adjusted for confounding. The findings of our study also emphasize that some factors important to maternal depression might change over the course of child rearing. From a policy standpoint, women with a constellation of significant risk factors are at high risk for maternal depression and should be assessed for depression. For example, a community could develop interventions to minimize risks, such as educational and social support to depressive mothers after screening. Our results show that mothers who reported less than good health status were at a higher risk for maternal depression than mothers who reported good or excellent health. However, data did not suggest the significant relationship between general health and maternal depression in GEE analysis. For marital status, Nagy and Ungerer⁴² suggested that in families with a child with cystic fibrosis, mothers with supportive spouses were less likely to be depressed than mothers with nonsupportive spouses. It seems reasonable to suggest that a supportive family structure would be associated with fewer depressive symptoms in mothers. Although this finding was not statistically significant in our study, this may be because of our inability to adequately measure this factor.

Our study has several strengths. A large sample of mothers was recruited from different places across the United States, which should increase the external validity. Maternal depression status was assessed at different times after birth of the child and, therefore, allowed evaluation of different time periods during child rearing. Because of the longitudinal follow-up, the prevalence of maternal depression over time and the pattern of change and persistence in risks could be evaluated, providing a description of the dynamic process of maternal depression. Furthermore, multivariate logistic regression and GEE analysis were used to provide simultaneous control for confounding and evaluation of risk for the different factors or variables. We believe the study provides a useful addition to the current literature on maternal depression.

Our study also has several limitations. The initial data were collected decades ago, and the results may not accurately represent the current situation regarding birth, child rearing, and the risk of maternal depression. People may be highly sensitive to questions about depression; therefore, there is some risk of self-reporting bias. In addition, the measure used (the CES-D) in this study is a screening tool rather than a diagnostic tool. The CES-D may not be as accurate as clinical diagnosis in women who are in the early postpartum weeks and months as it is at 1, 2, and 3 years after birth. The CES-D does not consider the disruptions to sleep, appetite, and energy that are normally present in early postpartum depression. Thus, the CES-D scores for these early times may not accurately measure maternal depression. Mothers who were not sufficiently conversant in English were excluded from this study, which prevented a more comprehensive health disparities analysis. For practical reasons, however, including only those who spoke English minimized some interpretation differences that would have occurred. Although a number of parental, family, social, and child variables were analyzed and adjusted for confounding in this study, there were additional potentially important variables, such as depression treatment, that were not included because of lack of pertinent data information.

Footnotes

Disclosure Statement

No competing financial interests exist.

References

Murray

CJL

, Lopez

. The global burden of disease: A comprehensive assessment of mortality and disability from diseases, injuries, and risk factors in 1990 and projected to 2020. Cambridge, MA: Harvard University Press, 1996.

Kessler

. Epidemiology of women and depression. J Affect Disord, 2003; 74:5–13.

World Health Organization. The World Health Report 2001, Mental health: New understanding, new hope. Geneva: World Health Organization.

Jiang

, Elixhauser

, Nicholas

, Steiner

, Reyes

, Bierman

. ‘Care of women in U.S. hospitals, 2000. HCUP fact book No.3, AHRQ Publication No. 02-004. Rockville, MD: Agency for Healthcare Research and Quality, 2002.

Kessler

, Berglund

, Demler

et al.

The epidemiology of major depressive disorder: Results from the National Comorbidity Survey Replication (NCS-R)

JAMA, 2003; 289:3095–3105.

Gaynes

, Gavin

, Meltzer-Brody

et al. Perinatal depression: Prevalence, screening accuracy, and screening outcomes. Evidence report/technology assessment No. 119. AHRQ Publication No. 05-E006–2. Rockville, MD: Agency for Healthcare Research and Quality, 2005.

Gavin

, Gaynes

, Lohr

, Meltzer-Brody

, Gartlehner

, Swinson

. Perinatal depression: A systematic review of prevalence and incidence. Obstet Gynecol, 2005; 106:1071–1083.

Chaudron

, Szilagyi

, Kitzman

, Wadkins

, Conwell

. Detection of postpartum depressive symptoms by screening at well-child visits. Pediatrics, 2004; 113:551–558.

Weissman

, Feder

, Pilowsky

et al. Depressed mothers coming to primary care: Maternal reports of problems with their children. J Aff Disord, 2004; 78:93–100.

10.

Rahman

, Harrington

, Bunn

. Can maternal depression increase infant risk of illness and growth impairment in developing countries? Child Care Health Dev, 2002; 28:51–56.

11.

World Health Organization. Mental health aspects of women's reproductive health: A global review of the literature. Geneva: World Health Organization, 2009.

12.

Weissman

, Pilowsky

, Wickramaratne

et al. Remissions in maternal depression and child psychopathology. JAMA, 2006; 295:1389–1398.

13.

Cohen

, Wills

. Stress, social support and the buffering hypothesis. Psychological Bulletin, 1985; 98:310–357.

14.

Weiss

. The provisions of social relationships. Rubin

. Doing unto others. Englewood Cliffs, NJ: Prentice Hall, 1974.

15.

Radloff

. The CES-D scale: A self-report depression scale for research in the general population. Appl Psychol Meas, 1977; 1:385–401.

16.

McDowell

, Newell

. Measuring health. A guide to rating scales and questionnaires. New York: Oxford University Press, 1996.

17.

Myers

, Weissman

. Use of a self-report symptom scale to detect depression in a community sample. Am J Psychiatry, 1980; 137:1081–1084.

18.

Cox

, Holden

, Sagovsky

. Detection of postnatal depression. Development of the 10 item Edinburgh Postnatal Depression Scale. Br J Psychiatry, 1987; 150:782–786.

19.

Evans

, Heron

, Francomb

, Oke

, Golding

. Cohort study of depressed mood during pregnancy and after childbirth. BMJ, 2001; 323:257–260.

20.

Dietz

, William

, Callaghan

, Bachman

, Whitlock

, Hornbrook

. Clinically identified maternal depression before, during and after pregnancies ending in live births. Am J Psychiatry, 2007; 164:1515–1520.

21.

Wellings

, Wadsworth

, Johnson

, Field

, Macdowall

. Teenage fertility and life chances. Rev Reprod, 1999; 4:184–190.

22.

Berndt

, Koran

, Finkelstein

et al. Lost human capital from early-onset chronic depression. Am J Psychiatry, 2000; 157:940–947.

23.

Onozawa

, Kumar

, Adams

, Dore

, Glover

. High EPDS scores in women from ethnic minorities living in London. Arch Womens Ment Health, 2003; 6:51–55.

24.

Peterson

, Hawley

. Effects of stressors on parenting attitudes and family function in a primary prevention program. Fam Relations, 1998; 47:221–227.

25.

Barnet

, Joffe

, Duggan

, Wilson

, Repke

. Depressive symptoms, stress, and social support in pregnant and postpartum adolescents. Arch Pediatr Adolesc Med, 1996; 150:64–69.

26.

Cairney

, Thorpe

, Rietschlim

, Avison

. 12-Month prevalence of depression among single and married mothers in the 1994 National Population Health Survey. Can J Public Health, 1999; 90:320–324.

27.

Lipman

, Offord

, Boyle

. Single mothers in Ontario: Sociodemographic, physical and mental health characteristics. Can Med Assoc J, 1997; 156:639–645.

28.

Wang

. The difference between single and married mothers in the 12 month prevalence of major depressive syndrome, associated factors and mental health service utilization. Soc Psychiatry Psychiatr Epidemiol, 2004; 39:26–32.

29.

Ali

, Avison

. Employment transitions and psychological distress: The contrasting experiences of single and married mothers. J Health Soc Behav, 1997; 38:345–362.

30.

Avison

. Family structure and mental health. Maney

, Ramos

. Socioeconomic conditions, stress and mental disorders: Toward a new synthesis of research and public policy. Bethesda, MD: NIH Office of Behavioral and Social Research, 2002; 1–38.

31.

Gyamfi

, Brooks-Gunn

, Jackson

. Association between employment and financial and parental stress in low income single black mothers. Welfare Work Well-Being, 2001; 32:119–135.

32.

Perez

, Beaudet

. The health of lone mothers. Health Rep, 1999; 11:21–32.

33.

Judge

. Parental coping strategies and strengths in families of young children with disabilities. Fam Relations, 1998; 47:263–268.

34.

Pietrukowicz

MCLC

. Apoio social e religião: Uma forma de enfrentamento dos problemas de saúde. [Mestrado] Rio de Janeiro: Fundação Oswaldo Cruz, Escola Nacional de Saúde Pública, 2001.

35.

Panzarine

, Slater

, Sharps

. Coping, social support, and depressive symptoms in adolescent mothers. J Adolesc Health, 1995; 17:113–119.

36.

Hudson

, Elek

, Campbell-Grossman

. Depression, self-esteem, loneliness, and social support among adolescent mothers participating in the new parents project. Adolescence, 2000; 35:445–453.

37.

Reid

, Meadows-Olive

. Postpartum depression in adolescent mothers: An integrative review of the literature. J Pediatr Health Care, 2007; 21:289–298.

38.

Liabsuetrakul

, Vittayanont

, Pitanupong

. Clinical applications of anxiety, social support, stressors, and self-esteem measured during pregnancy and postpartum for screening postpartum depression in Thai women. J Obstet Gynaecol Res, 2007; 33:333–340.

39.

Logsdon

, Birkimer

, Simpson

, Looney

. Postpartum depression and social support in adolescents. J Obstet Gynecol Neonatal Nurs, 2005; 34:46–54.

40.

Caldwell

, Antonucci

. Perceptions of parental support and depressive symptomatology among black and white adolescent mothers. J Emot Behav Disord, 1997; 5:173–183.

41.

Birkeland

, Thompson

, Phares

. Adolescent motherhood and postpartum depression. J Clin Child Adolesc Psychol, 2005; 34:292–300.

42.

Nagy

, Ungerer

. The adaptation of mothers and fathers to children with cystic fibrosis: A comparison. Child Health Care, 1990; 29:147–154.