PostPartum Depression in African American Women I need a paper written on POSTPARTUM DEPRESSION IN AFRICAN AMERICAN WOMEN. Paper needs to be in APA format, with an abstract. I will provide all ARTICLES. Please DO NOT use outside articles. ALL articles u

P1: GRA

Maternal and Child Health Journal (MACI) pp945-maci-470713 August 9, 2003 13:39 Style file version June 22, 2002

Maternal and Child Health Journal, Vol. 7, No. 3, September 2003 ( C© 2003)

Prevalence, Stability, and Socio-Demographic Correlates of Depressive Symptoms in Black Mothers During the First 18 Months Postpartum

Marjorie Beeghly, PhD,1,2 Karen L. Olson, PhD,1,2,3 M. Katherine Weinberg, PhD,1,2 Snaltze Charlot Pierre, BA,2 Nikora Downey, BA,2 and Edward Z. Tronick, PhD1,2,3

Objectives: The goals of this longitudinal study were to evaluate 1) the prevalence and stability of high depressive symptom levels during the first 18 months postpartum in a sample of otherwise healthy Black mothers varying in socio-economic status and 2) the relation of socio- demographic variables and level of socio-demographic risk to maternal depressive symptom levels during this time period. Methods: Participants were 163 Black adult mothers of healthy, full-term infants. The level of mothers’ depressive symptomatology was assessed at 2, 3, 6, 12, and 18 months postpartum using the Center for Epidemiological Studies-Depression Scale (CES-D). Mothers provided socio-demographic information at each assessment. Univariate and bivariate analyses were used to analyze the data. Results: The percentage of mothers with an elevated CES-D score (16 or higher) at single visits ranged from 13.5 to 14.7%, and 35.0% had at least one elevated CES-D score by 18 months postpartum. CES-D total scores were significantly correlated across each pair of visits (mean r = 0.57, all p’s < 0.0001), and average CES-D scores did not change significantly over time. Single marital status, low-income status, and more negative maternal perceptions of the adequacy of income for meeting familial needs were significantly related to higher maternal CES-D scores at each assessment (all p’s < 0.05). Level of socio-demographic risk, as assessed with a composite risk score derived from these variables, was significantly related to higher average CES-D scores (averaged across visits) ( p < 0.0001) and to a greater frequency of elevated CES-D scores (16 or higher) during the first 18 months postpartum ( p = 0.0002). Conclusions: The prevalence and stability of high levels of maternal depressive symptomatology during the first 18 months postpartum in this sample of Black women are consistent with those reported in prior studies of community samples of mothers unselected for race. Mothers with higher socio-demographic risk profiles had higher levels of maternal depressive symptoms at each assessment point.

KEY WORDS: maternal depression; Center for Epidemiological Studies-Depression Scale (CES-D); African Americans/Blacks; prevalence; stability; socio-demographic risk.

1 Department of Pediatrics, Harvard Medical School, Boston, Massachusetts.

2 Child Development Unit, Division of General Pediatrics, Chil- dren’s Hospital, Boston, Massachusetts.

3 Harvard University School of Public Health, Boston, Massachusetts.

4 Correspondence should be addressed to Marjorie Beeghly, Ph.D., Child Development Unit, Children’s Hospital, 1295 Boylston Street, Suite 320, Boston, MA 02215; e-mail: [email protected].

INTRODUCTION

Maternal postpartum depression is a prevalent and serious public health problem. In community samples, the prevalence of new mothers reporting clinically significant levels of depressive symptoms beyond the immediate postpartum period is ap- proximately 8–18% (1–10). For mothers with high socio-demographic risk profiles (e.g., low income, low education, single marital status, and/or young

157

1092-7875/03/0900-0157/0 C© 2003 Plenum Publishing Corporation

P1: GRA

Maternal and Child Health Journal (MACI) pp945-maci-470713 August 9, 2003 13:39 Style file version June 22, 2002

158 Beeghly, Olson, Weinberg, Pierre, Downey, and Tronick

maternal age), the prevalence is substantially higher (11–15), with percentages of 35% to 67% reported in some studies (16–19).

High levels of maternal depressive symptoms are of concern to health professionals because they are as- sociated with poor maternal psychosocial adaptation, compromised parenting, and child socio-emotional dysfunction, including insecure attachment, higher rates of behavior problems, aggression, and later psy- chiatric problems (20–25). Moreover, these associa- tions are stronger when maternal depressive symptom levels are severe or chronic (16, 26–30).

Several investigators have reported that women of color (particularly Black women) have higher lev- els of depressive symptomatology than White women (28, 31). However, this claim is difficult to evaluate because Black race/ethnicity and low socio-economic status (SES) are often confounded in this literature (32). Mothers from low-SES backgrounds in the U.S., who disproportionately tend to be women of color, typically report higher levels of depressive symptoms than their higher SES counterparts (7, 19, 28, 33–37). This has made it difficult to isolate the effects of de- pression from the effects of other risk factors and to disentangle the race-SES confounding often found in some prior studies of postpartum women. As a con- sequence, little is known about the prevalence and stability of high levels of depressive symptomatology in otherwise healthy postpartum Black women that is not confounded by these co-morbid risk factors.

Among studies that have controlled for socio- demographic variables, findings regarding racial differences in level of maternal depressive symptoma- tology have been inconsistent. In a sample of finan- cially impoverished, urban women (11), no racial dif- ferences in risk for depression were found, suggesting that poverty rather than race was the primary fac- tor influencing depression. Similarly, in a nationally representative sample of adolescent mothers from the National Maternal and Infant Survey (38), dif- ferences in depressive symptom levels between Black and White teen mothers decreased significantly after statistical control of family income and marital status. However, in a larger sample from the same national survey comprised of both teen and adult mothers (15), Black race continued to be associated with higher ma- ternal depressive symptom levels, even after control of several socio-demographic risk variables. Whether level of socio-demographic risk is related to higher levels of maternal depressive symptomatology within a single cohort of Black women ranging in socio- economic status has not been empirically established.

Moreover, little is known about the course of ma- ternal depressive symptomatology beyond the post- partum period, especially in non-White cohorts of women. In two large, racially diverse, nationally rep- resentative samples of women with young children (15, 16, 39), a moderate degree of stability in the level of mothers’ depressive symptoms was observed, as indicated by significant cross-time correlations in mothers’ scores on the Center for Epidemiological Studies-Depression Scale (CES-D) (40). Similar find- ings were reported in two smaller prospective studies of mostly White, middle-class women (41–44). It is un- clear, however, whether these findings generalize to otherwise healthy adult Black mothers from a range of socio-economic backgrounds.

The present study had two primary goals. The first was to evaluate the prevalence and stability of high levels of maternal depressive symptoms during the first 18 months postpartum in a selected commu- nity cohort of otherwise healthy Black adult mothers from a range of socio-economic backgrounds. Level of depressive symptomatology was assessed using the CES-D at 2, 3, 6, 12, and 18 months postpartum. A CES-D total score of 16 or higher was considered in- dicative of elevated depressive symptomatology. Four sets of questions associated with this first goal were addressed, as follows: 1) What percentage of moth- ers had a high CES-D score at each assessment? 2) Do mothers’ average CES-D total scores decrease over the course of the first 18 months postpartum, as reported in other studies (43, 44)? Alternatively, do their symptom levels increase or stay the same during this time period? 3) Of the women with an elevated CES-D score at intake (2 months postpartum), what percentage of women had another elevated CES-D score at later assessments? 4) Are mothers’ CES-D total scores at each visit significantly correlated over time (i.e., across each pair of visits)?

The second goal was to evaluate the relation be- tween socio-demographic risk variables and mothers’ depressive symptom levels during the first 18 months postpartum. Two questions associated with this sec- ond goal were addressed: 1) What is the relation be- tween individual socio-demographic variables (e.g., marital status, family income) and mothers’ CES- D scores at each assessment point (2, 3, 6, 12, and 18 months postpartum)? 2) Using a composite risk variable, does increasing levels of socio-demographic risk increase (a) the severity (average level) of mater- nal depressive symptoms during the first 18 months postpartum or (b) the persistence of high de- pressive symptom status (number of assessment

P1: GRA

Maternal and Child Health Journal (MACI) pp945-maci-470713 August 9, 2003 13:39 Style file version June 22, 2002

Depressive Symptoms in Black Mothers 159

periods at which a mother had an elevated CES-D total score)?

METHOD

Subjects

Subjects were participants in a larger, multi- hospital prospective longitudinal study of child de- velopment and parenting among Black adult mothers and their healthy, term infants. To be eligible for re- cruitment, mothers had to self-report as being African American or Black and be willing to provide informed consent. In addition, mother–infant dyads met a set of low-risk inclusion criteria. To be recruited, mothers had to be at least 18 years old and free of drug or alco- hol abuse problems, HIV positivity, and other serious chronic medical conditions, as confirmed by medical record review and maternal self-report. Infants had to be healthy, clinically normal at delivery, and full term (gestational weeks ≥37) as determined by med- ical record review and pediatric examination. Infants with obvious major congenital anomalies, a diagnosis of fetal alcohol syndrome, HIV positivity, or other se- rious health condition, or indication of prenatal drug exposure in the neonatal record were excluded.

Mothers and infants who met these inclusion cri- teria were recruited regardless of the infant’s gender, the mother’s marital status, parity, education, socio- economic status (SES), income, employment status, or family size. Maternal and infant characteristics at intake were obtained from a review of maternal and infant medical records and confirmed by infant physical examination at delivery and a follow-up ma- ternal interview at 2 months postpartum. These in- clusion criteria resulted in a medically low-risk yet demographically diverse sample.

Recruitment and Attrition

Recruitment of potential study mothers took place in two phases. Initial contact took place in the maternity wards of three metropolitan Harvard- affiliated teaching hospitals. After screening medi- cal records, a research assistant approached potential study mothers in the maternity ward and described the research protocol to them. Interested mothers com- pleted a brief demographic questionnaire and pro- vided signed informed consent to participate in a tele- phone interview when their baby was 2 months old.

Recruitment of eligible, consenting mothers into the follow-up phase of the study took place at the time of the 2-month telephone interview. Dyads who were re- cruited into the follow-up phase of the study were then observed four times during the child’s first 18 months: three times in the laboratory at 3, 6, and 18 months, and once in the child’s home at 12 months. Study mothers signed an additional informed consent form to participate in the follow-up phase of the study, at the time of the first laboratory visit. All recruitment and follow-up procedures received formal approval by the institutional human subjects review board.

A total of 794 mothers were approached in the maternity ward for potential recruitment. Of these, 679 (85.5%) mother–infant dyads met our inclusion criteria and were considered eligible to participate in the initial 2-month postpartum telephone inter- view. Two hundred sixty-eight (40%) mothers com- pleted the telephone interview and had their eligibil- ity status confirmed. Of the remaining mothers, 171 (25%) could not be reached, 157 (23%) were con- tacted at 2 months postpartum but were no longer interested in participating in the larger follow-up phase of the study, and 83 (12%) were contacted but no longer met inclusion criteria. There were no sig- nificant differences between the eligible, interested mothers who completed the 2-month telephone inter- view and those who did not on infant gender, maternal age (years), marital status (single/married), or parity (primiparous/multiparous).

Of the 268 eligible mothers who completed the 2-month phone interview, 195 mothers (73%) were recruited into the follow-up phase of the study and participated in at least one follow-up visit. There were no significant differences between this initial follow- up cohort of 195 mothers and the 73 nonparticipants on infant gender, infant birth weight (grams), num- ber of gestational weeks, or mothers’ primiparous versus multiparous status, or on total CES-D scores at intake. However, mothers who participated in the follow-up phase of the study were somewhat older, on average, than nonparticipating mothers (mean age of participants = 29.37, SD = 5.5; mean age of non-participants = 27.38, SD = 6.3, F [1, 266] = 6.34, p < 0.05) and had attained slightly more educa- tion (mean years for participants = 14.46, SD = 2.1; mean years for nonparticipants = 13.45, SD = 1.98, F [1, 259] = 11.49, p < 0.001).

Of the 195 mothers in the initial follow-up co- hort, 31 subjects (15.8%) subsequently dropped out of the study (21 moved or could not be reached, and 10 were not interested), and one missed the

P1: GRA

Maternal and Child Health Journal (MACI) pp945-maci-470713 August 9, 2003 13:39 Style file version June 22, 2002

160 Beeghly, Olson, Weinberg, Pierre, Downey, and Tronick

12-month study visit. Analyses in the present study were based on data from 163 mothers (83.6%) who finished the study and had complete data. Re- sults of chi square analyses indicated that these 163 mothers did not differ significantly from the 32 mothers who dropped out or had missing data on intake CES-D scores or on socio-demographic vari- ables, although participants were slightly more likely to have an income of less than $20,000 ( p = 0.093). Detailed data are available from the first author upon request.

Follow-up Assessments

Dyads were assessed in the laboratory at 3, 6, and 18 months postpartum and in their home at 12 months postpartum. At each follow-up visit, moth- ers completed the CES-D and other questionnaires and provided socio-demographic information about their employment status, occupation, child-care ar- rangements, and household composition (number of children and ratio of adults/children at home). At 3 and 18 months, mothers reported on their total yearly family income and their perceptions of the adequacy of their family’s total income for meeting familial needs. In addition, at each visit, mother–infant dyads were videotaped during social interaction.

Measurement of Variables

CES-D

The CES-D (40) is a widely used, self-report scale designed to measure level of depressive symptomatol- ogy in the general population. Possible scores range from 0 to 60, with higher scores reflecting higher levels of depressive symptoms in the past week. The CES- D has been shown to have excellent psychometric properties in both clinical and epidemiological stud- ies with diverse populations, including postpartum women (2, 40, 45). As is standard for this instru- ment and traditional in clinical studies of postpar- tum women (16, 44, 46, 47), a cutoff score of 16 or higher was used in this study as indicative of elevated depressive symptomatology.

Preliminary analyses were conducted to evalu- ate the internal consistency and concurrent validity of the CES-D in this cohort. Standardized Cronbach alpha coefficients for mothers’ CES-D total scores at each visit were all at or above 0.80 (0.82 at 2 months,

0.89 at 3 months, 0.87 at 6 months, 0.88 at 12 months, and 0.86 at 18 months). In support of the concur- rent validity of the CES-D, mothers’ CES-D scores at each assessment were significantly correlated with their concurrent scores on the Brief Symptom In- dex (BSI) (48), a widely used, abbreviated measure of maternal psychiatric symptomatology, including scores on its General Symptom Index (GSI) and De- pression Subscale (GSI: mean r [163] = 0.76, range = 0.68–0.83, all p’s < .0001; Depression subscale: mean r [163] = 0.72, range = 0.69–0.79, all p’s < 0.0001).

Socio-demographic Variables

A comprehensive list of socio-demographic vari- ables was evaluated as possible correlates of mater- nal CES-D scores, based on findings in the mater- nal depression literature, as follows: marital status (single/married), total family income (five-level cat- egorical variable), low-income status (yearly income < $20,000, which corresponds approximately to the federal poverty threshold for the majority of fami- lies in this sample, mothers’ perceptions of the ade- quacy of total income for meeting familial needs, as assessed using a five-point Likert rating (1 = very in- adequate, 5 = completely adequate), household com- position (the number of children and adults/children ratio in the household), maternal age (years), ma- ternal education (years), infant gender, infant birth weight (grams), infant gestational age (weeks), moth- ers’ country of origin (US-born/not), maternal em- ployment status (employed/not), number of hours per week the mother is employed, the number of hours per week that the infant was in nonparental child care, and SES. SES was assessed using the Hollingshead four-factor index of SES (49), which is based on the parent’s or parents’ level of educa- tion and occupational status and yields both a con- tinuous total score (higher scores reflect higher SES) and a categorical score (lower scores reflect higher SES).

Statistical Analysis

Univariate statistics were used to describe the sample’s characteristics. Bivariate statistics were used to evaluate the four questions associated with the first study goal, as follows: Chi square analyses were used to evaluate the percentage of mothers who had a high

P1: GRA

Maternal and Child Health Journal (MACI) pp945-maci-470713 August 9, 2003 13:39 Style file version June 22, 2002

Depressive Symptoms in Black Mothers 161

CES-D score at each assessment, and to compare the percentage of women with and without an elevated CES-D score at intake (2 months postpartum) who had another elevated CES-D score at later assess- ments. One-way repeated-measures analysis of vari- ance (ANOVA) was used to evaluate whether the average level of maternal depressive symptomatol- ogy (CES-D total score) for the sample as a whole changed over the course of the first 18 months post- partum. Pearson product–moment correlations were used to evaluate whether mothers’ CES-D total scores at each visit were significantly correlated over time (i.e., across each pair of visits).

To evaluate the two questions associated with the second study goal, the following analyses were used: One-way ANOVAs (for categorical variables) and Pearson product–moment correlations (for con- tinuous variables) were used to evaluate the relation of individual socio-demographic variables to moth- ers’ CES-D scores at each assessment point. Based on the results of these analyses, a composite risk variable was constructed to reflect individual mothers’ level of socio-demographic risk. Univariate statistics were used to calculate the number and percent of moth- ers at each risk level. Two summary CES-D depen- dent measures were also constructed, as follows: The arithmetic mean of mothers’ CES-D scores at 2, 3, 6, 12, and 18 months was calculated to represent the average level of maternal depressive symptoms dur- ing the first 18 months postpartum. To represent the persistence of high depressive symptom levels across the first 18 months postpartum, the number of assess- ment periods at which individual mothers had a high CES-D score was tallied (possible scores ranged from 0 to 5). Univariate statistics were used to calculate the mean and standard deviation of each summary CES-D dependent variable. Bivariate statistics were then used to calculate the mean and standard devia- tion (SD) of each dependent variable by level of socio- demographic risk. The significance of the relation between the socio-demographic risk composite vari- able and the two summary dependent measures was evaluated using one-way ANOVAs, followed by Tukey post-hoc tests. Findings were noted as statisti- cally significant where p ≤ 0.05.

RESULTS

Sample Characteristics

The sample consisted of 163 Black adult mothers and their healthy, term infants (49.7% male, 44.8%

first-born) from heterogeneous socio-economic back- grounds. At intake, mothers varied in marital status (41.7% were single), household composition (num- ber of children at home ranged from 1 to 5), family income (23% had a total income of less than $20,000), and employment status (50.6% were employed). Re- gardless of their employment status, the majority of mothers reported that they were the infant’s pri- mary caregiver. Univariate statistics describing the maternal, infant, and socio-demographic characteris- tics of the sample are provided in Table I (first two columns).

Prevalence and Stability of Maternal Depressive Symptomatology

Descriptive statistics for the number and percent- age of mothers with a high CES-D score at each as- sessment period, and the number of visits at which a mother received an elevated CES-D score, are pro- vided in Table II. The percentage of mothers with a high CES-D score at single visits ranged from 13.5% to 14.7%, and the number of mothers with at least one high CES-D score increased over time. By 18 months postpartum, 35.0% of mothers had at least one high CES-D score.

Descriptive statistics for mothers’ mean CES-D total score at each assessment period and mothers’ mean CES-D total score, averaged across visits, are provided in Table II. Results of repeated-measures ANOVA indicated that mothers’ mean total CES-D score at each visit did not change significantly over time.

In Table III, a breakdown is provided of the per- centage of mothers with a high (16+) or a lower (0–15) CES-D score at 2 months postpartum (intake) who had a high CES-D score at later assessment points (i.e., at 3, 6, 12, and 18 months postpartum), along with results of chi square analyses, odds ratios, and 95% confidence intervals (CIs). Results indicated that mothers with a high CES-D score at 2 months postpar- tum were significantly more likely than their lower- scoring counterparts to have another high CES-D score later during the first 18 months postpartum. Over a third of mothers with a high CES-D score at intake had another elevated CES-D score at each sub- sequent assessment.

Results of correlational analyses indicated that maternal CES-D scores were significantly and substantially correlated across all visits (mean r [163] = 0.57, range = 0.47 to 0.66, all

P1: GRA

Maternal and Child Health Journal (MACI) pp945-maci-470713 August 9, 2003 13:39 Style file version June 22, 2002

162 Beeghly, Olson, Weinberg, Pierre, Downey, and Tronick

Table I. Sample Characteristics at Intake

Maternal depressive symptom status at 2 months postpartum (intake)

High CES-D score Lower CES-D Total sample (16 or higher) score (0–15)

N = 163 (n = 23) (n = 140) Continuous variables Mean SDa Mean SD Mean SD F (1,161) p value

Number of children at home 1.97 1.04 2.30 1.40 1.91 0.097 2.78 0.0972 Ratio adults to children at home 1.45 0.84 1.34 0.89 1.47 0.83 0.44 0.5075 Hollingshead SESb index 43.69 11.56 44.46 10.89 43.56 11.70 0.12 0.7328 Perceived adequacy of income for meeting 3.17 1.03 2.64 1.20 3.26 0.98 7.52 0.0068

family needsc

Hours per week mother is employedd 15.85 18.14 22.35 18.92 14.78 17.85 3.49 0.0634 Hours per week infant receives 14.43 19.14 20.65 21.44 13.52 18.63 2.77 0.0979

nonparental child caree

Maternal age (years) 29.58 5.46 29.14 6.20 29.65 5.35 0.17 0.6793 Maternal education (years) 14.55 2.13 14.43 2.23 14.56 2.12 0.07 0.7881 Infant birth weight (grams) 3463 514 3416.0 558.1 3470.5 508.2 0.22 0.6386 Infant gestational age (weeks) 39.6 1.3 39.78 1.00 39.51 1.32 0.91 0.3413

Categorical variables Number Percent Number Percent Number Percent χ2 df p value

Infant gender Male 81 49.7 13 56.5 68 48.6 0.50 1 0.4797 Female 82 50.3 10 43.5 72 51.4

Parity Primiparous 90 65.2 9 39.1 64 45.7 0.35 1 0.5562 Multiparous 73 44.8 14 60.9 76 54.3

Mother is US born No 32 19.6 3 13.0 29 20.7 0.74 1 0.3907 Yes 131 80.4 20 86.9 111 79.3

Marital status Single 68 41.7 15 65.2 53 37.9 6.08 1 0.0137 Married 95 58.3 8 34.8 87 62.1

Mother is infant’s primary caretaker No 30 18.9 5 21.7 25 18.4 0.14 1 0.7035 Yes 129 81.1 18 78.3 111 81.6

Mother is employed No 79 49.4 9 39.1 70 51.1 1.13 1 0.2882 Yes 81 50.6 14 60.9 67 48.9

Yearly total income > $100,000 8 4.9 1 4.4 7 5.0 6.91 6 0.3295 $75–100,000 27 16.6 3 13.0 24 17.1 $50–75,000 29 17.8 3 13.0 26 18.6 $30–50,000 38 23.3 4 17.4 34 24.3 $20–30,000 23 14.1 4 13.4 19 13.6 $10–20,000 23 14.1 7 30.4 16 11.4 <$10,000 15 9.2 1 4.4 14 10.0

Low-income status f

No 125 79.7 15 65.2 110 78.6 1.97 1 0.1604 Yes 38 23.3 8 34.8 30 21.4

a SD = standard deviation. bSES = socio-economic status. c Five-point Likert scale, higher scores reflect greater adequacy. d Includes unemployed mothers, who received a score of 0. Among employed mothers, the mean number of work hours was 34.4 ± 8.4. e Includes infants not receiving nonparental childcare, who received a score of 0. Among infants receiving nonparental childcare, the mean number of child care hours was 34.3 ± 13.5.

f Yearly income <$20,000.

P1: GRA

Maternal and Child Health Journal (MACI) pp945-maci-470713 August 9, 2003 13:39 Style file version June 22, 2002

Depressive Symptoms in Black Mothers 163

Table II. Descriptive Statistics for CES-D Variables at Each Assessment Period

Higha CES-D score Average CES-D score

Assessment period Number Percent Mean SDb

2 months postpartum 23 14.1 8.06 6.95 3 months postpartum 24 14.7 8.84 8.14 6 months postpartum 22 13.5 8.14 7.43 12 months postpartum 23 14.1 7.99 7.36 18 months postpartum 24 14.7 8.94 7.01

Summary variables Mean CES-D total scorec 8.40 6.02 Mean number of high CES-D scores across all visits 0.71 1.20 Any higha CES-D score by 12 months 52 31.9 Any higha CES-D score by 18 months 57 35.0

Note. Results of a one-way repeated measures ANOVA revealed no significant change in average CES-D scores by time of assessment (F [4, 159] = 1.59, p = .1786). a High CES-D score refers to a score of 16 or higher. bSD = standard deviation. c Mean of CES-D total scores at 2, 3, 6, 12, 18 months postpartum.

p’s < 0.0001). Specific results are provided in Table IV.

Relation of Socio-demographic Variables to CES-D Scores at 2, 3, 6, 12, and 18 Months Postpartum

In the six right-hand columns of Table I, a break- down of the sample’s socio-demographic character- istics by mothers’ high (16+) versus lower (0–15) CES-D total scores at 2 months postpartum (intake) is presented. Results of bivariate analyses indicated that two of the 18 socio-demographic variables (sin- gle marital status and negative maternal perceptions of the adequacy of income for meeting familial needs) were significantly related to high versus lower mater- nal depressive symptom status at intake. Four addi- tional socio-demographic variables were marginally related ( p < 0.10) to high maternal depressive symp- tom status at intake: number of children at home ( p = 0.097), the ratio of the number of adults to the

Table III. Number and Percent of Mothers With a Higha CES-D Score at Intake (2 Months Postpartum) Who Also Had a Higha CES-D Score at Later Assessments

Number and percent of mothers with a higha CES-D score

3 months 6 months 12 months 18 months 2 month CES-D

score N % N % N % N %

High (16+), N = 23 11 47.8 10 43.5 8 34.8 8 34.8 Lower (0-15), N = 140 13 9.3 12 8.6 15 10.7 16 11.4 χ2 (df = 1) 23.37, p < .0001 20.62, p < .0001 9.44, p = .0021 8.58, p = .0034 Odds ratio (95% CIb) 8.96 (3.3, 24.3) 8.21 (3.0, 22.6) 4.44 (1.6, 12.2) 4.13 (1.5, 11.3) a High CES-D score refers to a score of 16 or higher. bCI = confidence interval.

number of children at home ( p = 0.058), a greater number of hours per week the mother was working ( p = 0.063), and a greater number of hours per week that the infant was in non-parental care ( p = 0.098).

Additional bivariate analyses were conducted to evaluate the relation of these 18 socio-demographic variables to maternal CES-D total scores at each as- sessment period. Results indicated that three of the 18 socio-demographic variables (single marital status, low-income status, and negative perceptions of the adequacy of income for meeting familial needs) were significantly related to maternal CES-D total scores at each assessment period.

Three additional variables (low SES, more than four children at home, and infant gestational age) were significantly associated with maternal total CES-D scores, but only at one assessment point. At 6 months postpartum, mothers with a low SES score or four or more children at home had a higher CES-D score than their lower-risk counterparts. For mothers

P1: GRA

Maternal and Child Health Journal (MACI) pp945-maci-470713 August 9, 2003 13:39 Style file version June 22, 2002

164 Beeghly, Olson, Weinberg, Pierre, Downey, and Tronick

Table IV. Correlations of CES-D Scores Over Time

Assessment perioda

2 months 3 months 6 months 12 months

3 months 0.62 6 months 0.52 0.66 12 months 0.48 0.60 0.65 18 months 0.47 0.59 0.61 0.56

Note. N = 163. All correlations are significant, p < .0001. a Number of months postpartum.

with low SES, the mean CES-D score was 11.53 (SD = 10.22), compared to a mean CES-D of 7.69 (SD = 6.90) for mothers without low-income status, F (1, 161) = 4.57, p = 0.034. For mothers with four or more children at home, the mean CES-D score was 12.71 (SD = 11.93), compared to a mean CES-D score of 7.71 (SD = 6.76), for mothers with fewer than four children at home, F (1, 161) = 5.99, p = 0.0155. At the 12-month assessment period, higher infant gestational age (weeks) was significantly, positively correlated with maternal CES-D scores (r [163] = 0.15, p < 0.05). All other socio-demographic vari- ables were not significantly related to maternal CES- D total scores at any assessment period.

In Table V, descriptive statistics and results of one-way ANOVAs are presented for the three

Table V. The Association of Selected Socio-demographic Risk Variablesa With Mothers’ Mean CES-D Total Scores at Each Assessment Period

Maternal CES-D total score

2 months 3 months 6 months 12 months 18 months

N Mean SDb Mean SD Mean SD Mean SD Mean SD

Single marital status No 95 6.77 6.55 7.22 6.96 7.00 7.77 6.11 5.87 7.79 7.06 Yes 68 9.87 7.14 11.11 9.12 9.74 6.65 10.62 8.39 10.56 6.65

F (1, 161) = 8.22, 9.57, 5.53, 16.3, 6.40, p = 0.0047 p = 0.0023 p = 0.0199 p < 0.0001 p = 0.0124

Low-income statusc No 125 7.47 6.78 7.61 6.82 7.41 7.52 6.86 6.90 7.89 6.80 Yes 38 10.00 7.26 12.88 10.60 10.55 6.66 11.68 7.68 12.42 6.61

F (1, 161) = 3.92, 13.1, 5.37, 13.5, 13.1, p = 0.0494 p = 0.0004 p = 0.0218 p = 0.0003 p = 0.0004

Negative maternal No 137 6.85 5.97 7.07 5.67 7.29 6.93 6.70 6.63 8.27 6.71 perceptions of the adequacy Yes 26 14.42 8.33 18.17 12.11 12.62 8.45 14.77 7.40 12.50 7.57 of total income for meeting F (1, 161) = 30.6, 54.0, 12.0, 31.2, 8.33, familial needsd p < 0.0001 p < 0.0001 p = 0.0007 p < 0.0001 p = 0.0044

Note. Results of one-way ANOVAs comparing the CES-D scores of mothers with and without each risk variable were significant at all assessment periods. a These variables were significantly associated with mothers’ CES-D total scores at each assessment period. bSD = standard deviation. c Yearly income <$20,000. d Score of 1 or 2 on a five-point Likert rating.

risk variables that were significantly related to mother’s total CES-D scores at each visit. The socio-demographic variables in this table were col- lected at the 2-month intake point. However, re- sults were virtually identical whether these risk vari- ables were derived at the 2-month intake point or at later assessments. Complete data on all risk variables are available upon request from the first author.

Composite Socio-demographic Risk Variable

Based on these findings, a composite risk vari- able was constructed to assess individual mothers’ level of socio-demographic risk. One risk point was as- signed for each of the following: single marital status, low-income status (total family income < $20,000), and negative maternal perceptions of the adequacy of total family income for meeting familial needs (rat- ing of 1 or 2 on the five-point Likert scale). Risk points were then summed for each subject (possible scores ranged from 0 to 3). Eighty mothers (49.08%) had 0 risk points, 46 (28.2%) had 1 risk point, 25 (15.34%) had 2 risk points, and 12 (7.36%) had 3 risk points.

P1: GRA

Maternal and Child Health Journal (MACI) pp945-maci-470713 August 9, 2003 13:39 Style file version June 22, 2002

Depressive Symptoms in Black Mothers 165

Relation of Level of Socio-Demographic Risk to Summary CES-D Variables

In Table II, univariate descriptive statistics are provided for the two summary CES-D variables used in these analyses: mothers’ mean CES-D score (aver- aged across visits) and the number of visits at which a mother had a high CES-D score. In Table VI, de- scriptive statistics for these two summary variables are presented, as broken down by the four-level com- posite risk variable.

Results of one-way ANOVAs indicated that the composite risk variable was significantly associated with mothers’ averaged CES-D score (F [3, 159] = 14.56, p < 0.0001). Results of Tukey post hoc tests indicated that mothers with 3 risk points had a signif- icantly higher averaged CES-D score than mothers with either 2, 1, or 0 risk points. In turn, mothers with 1 risk point had a significantly higher averaged CES-D score than mothers with 0 risk points.

A similar pattern of results was obtained for the persistence variable (F [3, 159] = 7.02, p = 0.0002). Results of Tukey post hoc tests indicated that mothers with 3 risk points had a significantly greater number of high CES-D scores during the first 18 months post- partum than mothers with either 2, 1, or 0 risk points. Similarly, mothers with 2 risk points had a significantly greater number of high CES-D scores than mothers with 0 risk points.

DISCUSSION

This longitudinal study is among the first to eval- uate the prevalence and stability of high depressive

Table VI. Relation Between Level of Socio-demographic Risk and Summary CES-D Variables

Number of visits that Averaged a mother had a highb

CES-D total scorea CES-D score Risk score N Mean SDc Mean SD

0 80 6.09d 4.95 0.40d .092 1 46 9.48e 5.96 0.74de 1.24 2 25 10.09e f 4.49 1.12e 1.33 3 12 16.09g 7.11 1.83 f 1.53

F (3, 159) = 14.56, F (3, 159) = 7.02, p < 0.0001 p = 0.0002

Note. Means with different superscripts in each column were sig- nificantly different in Tukey post hoc analyses. a Arithmetic mean of mothers’ CES-D total scores at 2, 3, 6, 12, and 18 months postpartum.

bHigh CES-D score refers to a score of 16 or more. c SD = standard deviation.

symptom levels during the first 18 months postpartum in a selected community sample of otherwise healthy Black adult mothers of term infants. Mothers in this sample varied in socio-demographic characteristics, which made it possible for us to evaluate the rela- tion between varying levels of socio-demographic risk and the severity and chronicity of mothers’ depressive symptomatology during this time period.

Prevalence and Stability of Maternal Depressive Symptoms

Findings regarding the prevalence of high ma- ternal depressive symptomatology in this medically low-risk sample were largely consistent with those re- ported for mothers in low-risk community samples unselected for race (43, 44). In the present study, the prevalence of mothers with an elevated CES-D score (16 or higher) at single visits during the first 18 months postpartum ranged from 13.5% to 14.7%. By 18 months postpartum, over a third (35.0%) of mothers in the present study had at least one ele- vated CES-D score. The single-visit prevalence val- ues fall within the range of 8-18% reported in the maternal depression literature (1–10). These values are somewhat lower than those reported for moth- ers of toddlers and preschoolers in a more racially, economically, and medically diverse, national sample (15), in which 23.7% of mothers had a high CES-D score at 17 months postpartum, and 16.6% of moth- ers had a high CES-D score at 35 months postpartum. Notably, the percentage of mothers with at least one elevated CES-D score over the course of the national study (31.9%) was very similar to that observed in the present study (35%), although the specific time periods evaluated and the number of assessments in- cluded in each study differed. In contrast, the preva- lence of high levels of maternal depressive symptoms in the present study was significantly lower than that reported for mothers in uniformly high-risk samples, such as low-income or teen cohorts (11–15, 19, 38).

Results from this study also provide evidence for a substantial degree of stability in level of ma- ternal depressive symptomatology over the course of the first 18 months postpartum. Over a third of mothers who had a high CES-D score at 2 months postpartum had another high CES-D score at each subsequent assessment. Moreover, mothers’ CES-D scores were strongly correlated across each pair of as- sessment periods, suggesting that individual mothers retain their initial rank-ordering in level of depressive symptomatology over time. These results replicate

P1: GRA

Maternal and Child Health Journal (MACI) pp945-maci-470713 August 9, 2003 13:39 Style file version June 22, 2002

166 Beeghly, Olson, Weinberg, Pierre, Downey, and Tronick

and extend findings from prior research that evalu- ated the course of maternal depressive symptomatol- ogy during the first year postpartum in a low-risk co- hort of mostly White mothers (44). In that study, as in the present study, approximately a third of moth- ers with a high CES-D score at 2 months postpartum had another high CES-D score at 3, 6, and 12 months postpartum, and mothers’ CES-D scores were corre- lated across time. Similar findings were also reported in studies of racially and economically diverse cohorts of mothers of toddlers and preschoolers (15, 39).

It is noteworthy that, in the present study, mothers’ average CES-D scores at each assessment point did not diminish significantly during the first 18 months postpartum, in contrast to the significant remission reported in some other longitudinal stud- ies of medically low-risk postpartum women. In one set of studies (42, 43), only 13% of mothers who met diagnostic criteria for major or minor depression at 2 months postpartum continued to meet diagnostic criteria at 24 months postpartum. Nevertheless, ap- proximately a third of these women continued to re- port high levels of depressive symptoms by 24 months postpartum. Similarly, in other longitudinal research (44), mothers’ average CES-D score decreased sig- nificantly from 2 to 3 months postpartum but did not change from 3 to 12 months postpartum.

Socio-Demographic Risk

The findings in the present study do not support suggestions made in some prior reports that Black women living in the U.S. may universally have higher levels of depressive symptomatology than White women (28, 31). Rather, the mothers in the present study exhibited variability in the level of their de- pressive symptomatology during the first 18 months postpartum, which was associated with their socio- demographic risk status. In this sample, three specific socio-demographic risk variables (single marital sta- tus, low-income status, and negative maternal percep- tions of the adequacy of income for meeting familial needs) were associated with higher mean maternal CES-D scores at each assessment period. These find- ings are consistent with a large literature documenting that single mothers and low-income mothers tend to experience higher levels of depressive symptoms than their lower-risk counterparts (19, 36). More- over, as the number of these risk factors increased, mothers had increasingly higher levels of depressive symptomatology overall as well as a greater number of elevated CES-D scores during the first 18 months

postpartum. These results provide some general support for a cumulative risk hypothesis (50, 51).

In much of the maternal depression literature, Black race has been confounded with low SES. This is not surprising since Black families in the U.S. are disproportionately more likely than White families to live in poverty (33). However, a sizable number of Black families in the U.S. have higher socio-economic status. The results from this study suggest that further attention be given to this understudied group in future research.

Limitations

Several limitations of this study should be noted. Due to the study’s low-risk inclusion criteria, the findings of this study likely pertain only to healthy, adult Black mothers who are parenting healthy, full-term infants. Findings may not generalize to higher-risk groups of Black mothers, such as chron- ically ill mothers, teen mothers, mothers of sick or prematurely born infants, or mothers living in severe poverty. Moreover, the relatively small sample size (N = 163) may have limited statistical power needed to identify additional socio-demographic correlates of depression.

Another limitation is that the CES-D was the sole measure of maternal depressive symptomatology used in this study. Although the CES-D has sound psy- chometric properties and higher CES-D scores have been linked to a clinical diagnosis of depression in studies of postpartum women (44), and to poorer adaptive functioning, both in the general population (52–54) and among postpartum women (22, 47), the CES-D measures depressive symptom levels rather than depression per se. Thus, mothers with an ele- vated CES-D score do not necessarily have clinically significant depression.

For mother–infant dyads, higher levels of ma- ternal depressed mood and affect may be especially problematic, particularly if chronic (42), because they may create a negative “affective climate” for caregiv- ing that is likely to compromise the mother–infant re- lationship and ultimately the child’s functioning. For this reason, the substantial degree of stability in de- pressive symptom levels reported in this and other studies is clinically important and warrants further attention from public health professionals and health care practitioners. This is especially important be- cause primary care providers do not often recognize or treat maternal depression (55–57). Another obsta- cle to receiving care is that women with high levels

P1: GRA

Maternal and Child Health Journal (MACI) pp945-maci-470713 August 9, 2003 13:39 Style file version June 22, 2002

Depressive Symptoms in Black Mothers 167

of depressive symptoms (particularly if they are from low-SES backgrounds) may not always seek out pro- fessional help or discuss their symptoms with a health care professional (55). Because early identification of postnatally depressed mothers and early adequate treatment have been associated with a shorter dura- tion of postpartum depressive symptomatology (58), programs should be put into place in which health care providers routinely screen new mothers (regard- less of race) for depressive symptomatology early in the postpartum period, using the CES-D or similar screening instrument. This could take place at the in- fant’s pediatric visits or mothers’ visits to their own primary care provider. Mothers reporting high lev- els of depressive symptoms should be monitored or referred for further evaluation.

CONCLUSION

In conclusion, the results of this study indicate that the prevalence of high depressive symptom levels during the first 18 months postpartum among healthy, adult Black mothers varying in social class is largely consistent with that reported for mothers in commu- nity samples unselected for race. In the present study, mothers with higher socio-demographic risk profiles had significantly higher mean CES-D scores and a greater number of high CES-D scores during the first 18 months postpartum than their lower-risk counter- parts. Evidence for a substantial amount of stability in maternal depressive symptom levels over the course of the first 18 months postpartum was also found. Over a third of mothers with a high CES-D score at 2 months postpartum had another high CES-D score at subsequent assessments. Moreover, mothers’ CES-D scores were significantly correlated over time. These findings suggest that the high depressive symp- tom levels reported by some mothers early in the first postpartum year are not transient. Given that chronic depression can exert deleterious effects on maternal and child functioning (21, 47, 59), these findings sug- gest that further attention be given by public health professionals and primary care providers to the de- sign and implementation of depression screening pro- grams for postpartum women, regardless of race.

ACKNOWLEDGMENTS

The research presented in this paper was sup- ported in part by a Major Research Grant from the Spencer Foundation (Marjorie Beeghly, PI). The data

presented and views expressed are solely the respon- sibility of the authors. We thank the study families for their loyal participation in this longitudinal study and acknowledge the valuable research assistance of Caryn Brady, Lisa Fhagen, Holly Haynes, Jennifer Gray, Grace Gustafson, Jill Filene, Joan Driscoll, Re- becca Rosenberg, Adie Schulten, Jennifer Strickland, Iris Yates, and others at the Child Development Unit. This paper benefited from editorial feedback from Dr. Milton Kotelchuck and two anonymous reviewers.

REFERENCES

1. Weissman MM, Leaf, PJ, Tischler, GL, Blazer, DG. Affec- tive disorders in five United States communities. Psychol Med 1988;18(1):141–153.

2. Boyd JH, Weissman MM, Thompson WD, Myers JK. Screen- ing for depression in a community sample. Understanding the discrepancies between depression symptom and diagnostic scales. Arch Gen Psychiatry 1982;39:1195–1200.

3. Cooper PJ, Campbell EA, Day A, Kennerley H. Non-psychotic psychiatric disorder after childbirth: A prospective study of prevalence, incidence, course and nature. Br J Psychiatry 1988;152:799–806.

4. Murray L. The impact of postnatal depression on infant devel- opment. J Child Psychol Psychiatry 1992;33:543–561.

5. O’Hara MW, Swain AM. Rates and risk of postpartum depres- sion: A meta-analysis. Int Rev Psychiatry 1996;8:37–54.

6. O’Hara MW, Zekoski EM, Phillips LH, Wright EJ. Controlled prospective study of postpartum mood disorders: Comparison of childbearing and nonchildbearing women. J Abnorm Psy- chol 1990;99:3–15.

7. Beck CT. Predictors of postpartum depression: An update. Nurs Res 2001;50(5):275–285.

8. O’Hara MW. The nature of postpartum depressive disorders. In: Murray L, Cooper PJ, editors. Postpartum depression and child development. New York: Guilford Press 1997: 3–35.

9. Gotlib IH, Whiffen V, Mount J, Milne K, Cordy N. Prevalence rates and demographic characteristics associated with depres- sion in pregnancy and the postpartum. J Consult Clin Psychol 1989;57:269–274.

10. Cox JL, Murray D, Chapman G. A controlled study of the onset, duration, and prevalence of postnatal depression. Br J Psychiatry 1993;163:27–31.

11. Hobfoll SE, Ritter C, Lavin J, Hulsizer MR, Cameron RP. De- pression prevalence and incidence among inner-city pregnant and postpartum women. J Consult Clin Psychol 1995;63:445– 453.

12. Belle D. Poverty and women’s mental health. American Psy- chologist 1990;45(3):385–389.

13. Warner R, Appleby L, Whitton A, Faragher B. Demographic and obstetric risk factors for postnatal psychiatric morbidity. Br J Psychiatry 1996;168(5):607–611.

14. Blazer D, Kessler R, McGonagle K, Swartz M. The prevalence and distribution of major depression in a national community sample: The National Comorbidity Survey. Am J Psychiatry 1994;151:979–986.

15. McLennan JD, Kotelchuck M, Cho H. Prevalence, persistence, and correlates of depressive symptoms in a national sample of mothers of toddlers. J Am Acad Child Adolesc Psychiatry 2001;40:1316–1323.

16. Civic D, Holt VL. Maternal depressive symptoms and child behavior problems in a nationally representative normal birth- weight sample. Matern Child Health 2000;4(4):215–221.

P1: GRA

Maternal and Child Health Journal (MACI) pp945-maci-470713 August 9, 2003 13:39 Style file version June 22, 2002

168 Beeghly, Olson, Weinberg, Pierre, Downey, and Tronick

17. Colletta ND. At risk for depression: A study of young mothers. J Genet Psychol 1983;142:301–310.

18. Reis J. Correlates of depression according to maternal age. J Genet Psychol 1988;149:535–545.

19. Coiro MJ. Depressive symptoms among women receiving wel- fare. Women Health 2001;32(1–2):1–23.

20. Cummings EM, Davies PT. Maternal depression and child de- velopment. J Child Psychol Psychiatry 1994;35:73–112.

21. Downey G, Coyne JC. Children of depressed parents: An in- tegrative review. Psychol Bull 1990;108:50–76.

22. Field T. Psychologically depressed parents. In: Bornstein MH, editor. Handbook of parenting, Vol 4: Applied and practical parenting. Mahwah, NJ: Erlbaum, 1995: 85–99.

23. Tronick EZ, Weinberg MK. Depressed mothers and infants: Failure to form dyadic states of consciousness. In: Cooper PJ, editor. Postpartum depression and child development. New York: Guilford Press 1997: 54–81.

24. Hammen C. Children of affectively ill parents. In: Steinhausen HC, Verhulst F, editors. Risks and outcomes in developmental psychopathology. Oxford, UK: Oxford University Press, 1999: 38–53.

25. Kaslow NJ, Deering CG, Racusin GR. Depressed children and their families. Clin Psychol Rev 1994;14:35–59.

26. Teti DM, Gelfand DM, Pompa J. Depressed mothers’ behavioral competence with their infants: Demographic and psychosocial correlates. Dev Psychopathol 1990;2:259– 270.

27. Teti DM, Gelfand DM, Messinger DS, Isabella R. Maternal depression and the quality of early attachment: An examina- tion of infants, preschoolers, and their mothers. Dev Psychol 1995;31:364–376.

28. Garcia Coll C, Lamberty G, Jenkins R, McAdoo HP, Crnic KA, Wasik BH, Garcia HV. An integrative model for the study of developmental competencies in minority children. Child Dev 1996;67:1891–1914.

29. Brennan PA, Hammen C, Andersen MJ, Bor W, Najman JM, Williams GM. Chronicity, severity, and timing of maternal de- pressive symptoms: Relationships with child outcomes at age 5. Dev Psychol 2000;36:759–766.

30. Keller MB, Beardslee W, Dorer DJ, Lavori PW, Samuelson H, Klerman R. Impact of severity and chronicity of parental affective illness on adaptive functioning and psychopathology in children. Arch Gen Psychiatry 1986;43:930–937.

31. Garcia Coll C. Ethnic minority parenting. In: Bornstein MH, editor. Handbook of parenting Vol. 2: Biology and ecology of parenting. Mahwah, NJ: Erlbaum, 1995: 189–209.

32. Graham S. “Most of the subjects were white and middle class”: Trends in published research on African Americans in selected APA journals, 1970–1989. Am Psychol 1992;47(5):629–639.

33. National Center for Children in Poverty. Low-income children in the United States: A brief demographic profile. New York: Columbia University, 2002.

34. Corcoran M. Rags to rags: Poverty and mobility in the United States. Ann Rev Sociol 1995;21:237–267.

35. McLoyd VC, Wilson L. The strain of living poor: Parenting, social support, and child mental health. In: Huston AC, editor. Children in poverty: Child development and public policy, New York: Cambridge University Press, 1991: 105–135.

36. Huston AC, McLoyd VC, Garcia Coll C. Poverty and behavior: The case for multiple methods and levels of analysis. Dev Rev 1997;17:376–393.

37. Conger RD, Wallace LE, Sun Y, Simons RL, McLoyd VC, Brody GH. Economic pressure in African American families: A replication and extension of the Family Stress Model. Dev Psychol 2002;38:179–193.

38. Deal LW, Holt VL. Young maternal age and depressive symp- toms: Results from the 1988 National Maternal and Infant Health Survey. Am J Public Health 1998;88(2):266–270.

39. NICHD Early Child Care Research Network. Chronicity of maternal depressive symptoms, maternal sensitivity, and child functioning at 36 months. Dev Psychol 1999;35:1297–1310.

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

41. Campbell SB, Cohn JF, Flanagan C, Popper S, Meyers T. Course and correlates of postpartum depression during the transition to parenthood. Dev Psychopathol 1992;4:29– 47.

42. Campbell SB, Cohn JF, Meyers T. Depression in first-time mothers: Mother–infant interaction and depression chronic- ity. Dev Psychol 1995;31:349–357.

43. Campbell SB, Cohn JF. The timing and chronicity of post- partum depression: Implications for infant development. In: Murray L, Cooper PJ, editors. Postpartum depression and child development. New York: Guilford Press, 1997.

44. Beeghly M, Weinberg MK, Olson KL, Kernan H, Riley J, Tronick E. Stability and change in level of maternal depres- sive symptomatology during the first postpartum year. J Affect Disord 2002;71:169–180.

45. Husaini BA, Neff DA, Harrington JB, Hughes MD, Segal D. Depression in rural communities. Validating the CES-D scale. J Community Psychol 1980;8:20–27.

46. Campbell SB, Cohn JF. Prevalence and correlates of post- partum depression in first-time mothers. J Abnorm Psychol 1991;100:594–599.

47. Weinberg MK, Tronick EZ, Beeghly M, Olson KL, Kernan H, Riley J. Subsyndromal depressive symptoms and major depres- sion in postpartum women. Am J Orthopsychiatry 2001;71:87– 97.

48. Derogatis LR, Spencer R. The brief symptom inventory: Ad- ministration, scoring, and procedure manual. Baltimore, MD: Clinical Psychometric Research, 1982.

49. Hollingshead A. Four-factor index of social status. New Haven, CT: Yale University, 1979.

50. Seifer R, Sameroff AJ, Dickstein S, Keitner G, Miller I, Rasmussen S, et al. Parental psychopathology, multiple con- textual risks, and one-year outcomes in children. J Clin Child Psychol 1996;25:423–435.

51. Sameroff AJ, Seifer R, Baldwin A, Baldwin C. Stability of intelligence from preschool to adolescence: The influence of social and family risk factors. Child Dev 1993;64(1):80–97.

52. Judd LL, Martin PP, Wells KB, Rapaport MH. Socioeconomic burden of subsyndromal depressive symptoms and major de- pression in a sample of the general population. Am J Psychiatry 1996;153:1411–1417.

53. Judd LL, Akiskal HS, Zeller PJ, Paulus MP, Leon AC, Maser JD, et al. Psychosocial disability during the long-term course of unipolar major depressive disorder. Arch Gen Psychiatry 2000;57:375–380.

54. Gotlib IH, Lewinsohn PM, Seeley JR. Symptoms versus a di- agnosis of depression: Differences in psychosocial functioning. J Consult Clin Psychol 1995;63(1):90–100.

55. Scholle SH, Kelleher K. Preferences for depression ad- vice among low-income women. Maternal Child Health 2003;7(2):95–102.

56. Kelly RH, Zatzick DF, Anders TF. The detection and treatment of psychiatric disorder and substance use among pregnant women cared for in obstetrics. Am J Psychiatry 2001;158(2):213–219.

57. Davidson J, Meltzer-Brody S. The underrecognition and un- dertreatment of depression: What is the breadth and depth of the problem? J Clin Psychiatry 1999;60:S4–S9.

58. England SJ, Ballard C, George S. Chronicity in postnatal de- pression. Eur J Psychiatry 1994;8:93–96.

59. Cicchetti D, Toth SL. The development of depression in chil- dren and adolescents. Am Psychologist 1998;53(2):221–241.

Copyright of Maternal & Child Health Journal is the property of Springer Science & Business Media B.V. and

its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's

express written permission. However, users may print, download, or email articles for individual use.