Maternal and Child Health Journal

, Volume 11, Issue 3, pp 257–267

Depressive Symptomatology and Mental Health Help-Seeking Patterns of U.S.- and Foreign-Born Mothers

Authors

    • Department of International Health, School of Nursing and Health StudiesGeorgetown University
  • Frank Y. Wong
    • Department of International Health, School of Nursing and Health StudiesGeorgetown University
  • Cynthia R. Ronzio
    • Children’s National Medical Center
  • Stella M. Yu
    • Research & Demonstration BranchDRTE/MCHB/HRSA
Original paper

DOI: 10.1007/s10995-006-0168-x

Cite this article as:
Huang, Z.J., Wong, F.Y., Ronzio, C.R. et al. Matern Child Health J (2007) 11: 257. doi:10.1007/s10995-006-0168-x

Abstract

Objectives: This report presents the national estimates of maternal depressive symptomatology prevalence and its socio-demographic correlates among major racial/ethnic-nativity groups in the United States. We also examined the relationship of mental health-seeking patterns by race/ethnicity and nativity. Methods: Using the Early Childhood Longitudinal Survey-Birth Cohort Nine-month data, we present the distribution of Center for Epidemiological Study-Depression (CES-D) score by new mothers’ nativity and race/ethnicity. The mental health-seeking pattern study was limited to mothers with moderate to severe symptoms. Weighted prevalence and 95% confidence intervals for depression score categories were presented by race/ethnic groups and nativity. Multi-variable logistic regression was used to obtain the adjusted odds ratios of help-seeking patterns by race/ethnicity and nativity in mothers with moderate to severe symptoms. Results: Compared to foreign-born mothers, mothers born in the U.S. were more likely to have moderate to severe depressive symptoms in every racial/ethnic group except for Asian/Pacific Islanders. These US-born mothers were also more likely to be teenagers, lack a partner at home, and live in rural areas. Among Asians, Filipina mothers had the highest rate of severe depressive symptoms (9.6%), similar to those of US-born black mothers (10.2%). Racial/ethnic minorities and foreign-born mothers were less likely to consult doctors (OR: 2.2 to 2.5) or think they needed consultation (OR: 1.9 to 2.2) for their emotional problems compare to non-Hispanic White mothers. Conclusion: Our research suggests that previous “global estimates” on Asian American mental health underestimated sub-ethnic group differences. More efforts are needed to overcome the barriers in mental health services access and utilizations, especially in minority and foreign-born populations.

Keywords

Maternal depressionRacial/ethnic minorityForeign-bornCES-D

Introduction

According to the World Health Organization (WHO), depressive disorders are the leading cause of disability as measured by Years Lived with Disability and the fourth leading contributor to the global burden of disease in 2000 [1]. As one of the most prevalent diseases globally and an important cause of disability, depression is responsible for as many as one of every five visits to primary care doctors; it occurs universally and affects members of all racial/ethnic groups [2]. In the United States alone, the annual costs of depression totaled approximately 83.1 billion dollars in 2000, where 31% were direct medical costs, 7% were suicide-related mortality costs, and 62% were workplace costs [3].

Women are at higher risk for most types of depression [46], and depression is the leading cause of disease-related disability among women [7]. Approximately 12 million women in the U.S. experience depression every year—roughly twice the rate of men [8]. Pregnancy and new motherhood may increase the risk for the emergence of psychiatric illness, particularly in women who already have a history of mood disorders. Apart from inflicting distress on the mother, maternal depression undermines the marital relationship, impairs emotional and cognitive development of the newborn child [913], and may lead to postpartum psychosis [14]. Left untreated, one-third of the mothers continue to be depressed by the end of the postnatal year and one-tenth remain depressed by the end of the second postnatal year [15]. Although postnatal depression is common and potentially serious, only a minority of the cases are identified by primary care health workers during routine care [1618]. The majority of depressed mothers in the community are unrecognized and therefore untreated.

Despite the debilitating consequences of depression in pregnant women, the precise level of the prevalence and incidence of perinatal depression remains unknown [19]. Published estimates of the rate of major and minor depression in US women during the postpartum period range from 5% to more than 25% [2023]. There are also no published estimates of maternal depression by immigrant status and race/ethnic groups. Indeed, a recent evidence report published by the Agency for Healthcare Research and Quality (“Perinatal Depression: Prevalence, Screening Accuracy, and Screening Outcomes”) discusses the consequences of the lack of data on maternal depression among race/ethnic minorities:

The absence of information on populations other than the white population was dramatic. A better understanding of racial and ethnic variations could help clinicians know where to target screening programs and researchers know where to target studies on screening tools, and it could help researchers clarify the need for more nationally representative perinatal depression samples [19].

In addition, studies of maternal depression by race/ethnic group and associated risk factors may lead to a better understanding of the etiology of maternal depression and its cultural-specific components, which remain imperfectly understood. Common risk and protective factors include: social support, income, education, presence of a supportive spouse, history of abuse and violence exposure, and a history of depression or anxiety. Little is known about how the constellation of these factors actually affects the risk profile for maternal depression. Many of these factors likely possess a socio-cultural component in the role they play in mood disorders, and also may vary significantly with migration experience and by immigration status. It is useful to understand the role of immigration in the context of depression in two ways: One is the (possible) higher exposure rates to commonly accepted risk factors for maternal depression: and second is that: the psychological influence of the risk factors on maternal depression may vary socially and culturally.

Two such factors that may be particularly relevant for understanding the role of immigrant status are low social support and low socioeconomic status (SES) [2429]. For example, the relevance of low social support on maternal depression may vary by ethnic group, and the experience of low social support may also vary by immigration status and by the unique communal characteristics of immigrant groups in the US. The cohesiveness of the mother’s specific ethnic or national community within the US will affect social support.

Hobfoll characterizes the effects of low SES on postpartum depression as “the higher frequency of chronic stressful life events, less control over the occurrence of stressors and most compellingly, increased vulnerability to the negative effects of stressful events” [30]. Coupled with low SES, the experience of immigration as well as the status as an immigrant, which creates separation from long-standing sources of security and social support, may act as potentially chronic stressors. For example, it has been reported that Chinese, Filipino, Japanese, and Korean immigrants consistently report higher numbers of depressive symptoms than whites [31]. Meanwhile, Asian Americans in general reported the lowest utilization for mental health services and are more likely to present with severe mental health diagnoses such as psychoses at clinics. Studies further show that Asian Americans have greater disturbance levels than do non-Asian clients [32]. Attention to ethnicity and immigrant status is likely to shed light on maternal depression, especially among minority women living in the U.S.

This study was conducted using new data from the Early Childhood Longitudinal Study-Birth Cohort Nine-month interview data, which over-sampled various immigrant and ethnic groups. This is the first study to present national estimations of the burden of maternal depression and health-seeking behaviors by multiple racial/ethnic groups, with specific attention being paid to the role of nativity on rates of depressive symptomatology. The paper also describes the distributions of key risk factors for maternal depression by nativity and race/ethnic groups. While this current study cannot answer how the roles of these risk factors for maternal depression vary for women of different cultural backgrounds and with different immigration statuses, it is important to document their distribution for specific race/ethnic and immigrant groups.

Methods

Data source

The data used in this study are derived from the Early Childhood Longitudinal Study-Birth Cohort (ECLS-B) Nine-month survey conducted by the National Center for Education Statistics of the US Department of Education in collaboration with multiple centers from Department of Health and Human Services. The ECLS-B study is a multi-source, multi-method study that focuses on the early home and educational experiences of children during their first 6 years. It follows a nationally representative cohort of children born in 2001 from birth through first grade. The parents of 10,688 children born in 2001 participated in the first wave of the study when the children were approximately 9 months old. The base year data were collected on a rolling basis between the fall of 2001 and the fall of 2002. The central goal of the ECLS-B is to provide a comprehensive and reliable set of data that could be used to conduct in-depth analysis of children’s home and care experiences. There are a couple of aspects of the ECLS-B design that help to achieve this goal. One is over-sampling of American Indian, Asian and Pacific Islander infants as well as low birth weight infants; the other is the inclusion of a rich resource on parental mental and physical health as well as measures on the home environment and family functions. Thus, the ECLS-B provides us an unprecedented opportunity to study maternal depressive symptomatology, especially among ethnic minorities in the U.S.

The target population of the ECLS-B consists of all children born in the U.S. in the year 2001 except the following: (1) children born to mothers less than 15 years of age; (2) children who died before the 9-month assessment; and (3) children who were adopted prior to the 9-month assessment. Children were sampled primarily via registered births from the National Center for Health Statistics (NCHS) vital statistics system (U.S. DHHS, 1997). Its sample size was designed to produce survey estimates with specific precision targets both overall and for the following race/ethnicities: American Indian, Chinese, Other Asian or Pacific Islander, Hispanic, black non-Hispanic, or white non-Hispanic.

A number of steps were taken to improve response rates in the nine-month survey, such as use of incentives, the refusal conversion process, procedures for locating the sampled cases, etc. As a result, the overall response rate for the 9-month data collection was 74.1% [33]. All data collection instruments were translated into Spanish. For other languages, interpreters were used for most, but not all of the instruments. Spanish-speaking interviewers were recruited in the areas where Hispanic births were concentrated. For other languages, interviewers were instructed to use a member of the local community or a household member as an interpreter.

Because of the time that was required to locate families and to schedule and conduct the home visit, children ranged in age from approximately 6 months to 22 months at the time of the assessment in the ECLS-B nine-month data. For the present study we limited the analysis to children who were under 1 year and those mothers finished the self-administered questionnaire in order to capture maternal depression in more accurate terms. For analysis on help-seeking patterns, we used a sub-sample of mothers who had “moderate” or “severe” depression symptomatology based on their CES-D (Center for Epidemiologic Studies-Depression Scale) score.

Measures

Maternal depressive symptomatology and help-seeking patterns

The ECLS-B used a modified version of the CES-D which was validated in the Family and Child Experiences Survey (FACES) in 1999 [34]. Twelve questions were used in the parent self-administered questionnaire. We followed the specific instructions on ECLS-B data menu to calculate CES-D scales score based on the answers to these questions. According to the instructions, the CES-D was coded as missing for anyone who had more than 3 invalid (missing or refused) responses on the 12 items. Based on suggestions from the U.S. Department of Education we were able to categorize the symptoms as non-depressed (scores 0–4), mildly depressed (scores 5–9), moderately depressed (scores 10–14) and severely depressed (scores 15 or higher) [33]. Radloff recommended that respondents with a total CES-D score of 16 or higher should be screened for a diagnosis of major depression [35].

The help-seeking patterns were derived from the responses to the two questions: “In the past 12 months, have you talked with a psychiatrist, psychologist, doctor or counselor for any emotional or psychological problems?”; and “In the past 12 months, have you felt, or has anyone suggested, that you needed help for any emotional or psychological problems?”. The answers for both questions were limited to “yes” and “no.”

Maternal nativity and race/ethnicity

Mothers’ nativity was based on the mother’s place of birth (native born vs. foreign born) from the infant’s birth certificate. Information on race and ethnicity was collected during the home interview when the interviewer used flash cards and asked the mothers to select one or more of the racial and ethnic categories on the card that best describes their race. The available racial/ethnic groups include: White, Black, American Indian, Asian Indian, Chinese, Filipino, Japanese, Korean, Vietnamese, Other Asian, Native Hawai'ian, Guamian/Chamorro, Samoan, Other Pacific Islander, and Other. If the respondents answered “yes” to the question, “Are you of Spanish, Hispanic or Latino origin?”, they were also asked about which ethnic groups they belong to: Mexican, Puerto Rican, Cuban and Other Hispanic. In the present study, we collapsed some small groups and defined race/ethnicity as a 16 category variable which included non-Hispanic White, non-Hispanic Black, Mexican, Puerto Rican, Cuban, other Hispanic, and non-Hispanic Asian Indian, Chinese, Filipino, Japanese, Korean, Vietnamese, Other Asian, Pacific Islander, Native American, and Other.

Socio-demographic characteristics

Data on children’s age, sex, self-reported maternal physical health (excellent/very good/ good vs. fair/poor), annual family income (≤20K, 20–35K, 35–50K, 50–70K or 70k+), whether the mother had a partner at home (yes vs. no), parity (0 vs. 1+), and place of residence (urban, suburban or rural) were obtained from the at-home interview. A composite scale of family socio-economic status was available from ECLS-B dataset. It is a measure of social standing [36] and was computed at the household level using data from the parent’s computer-assisted personal interview (CAPI) instrument and the resident father questionnaire. The components used to create the measure included: (1) father/male guardian’s education and occupation, (2) mother/ female guardian’s education and occupation, and (3) household income. More details on this variable can be found elsewhere (User Manual, page 7–26 to 32). In our analysis the SES variable was a 5-category ordinal variable (1–5) based on the quintile of the SES scale.

Statistical analyses

SUDAAN software was used to generate all estimates and related standard errors to account for the complex sample design involving stratification, clustering and multistage sampling of the ECLS-B [37]. All estimates were weighted to reflect the families with children born in 2001 in the U.S. Chi-square statistics were used to test for differences in the proportion of degree of depressive symptomatology and mental health help-seeking patterns among racial/ethnic groups stratified by nativity. Distributions of SES, physical health, and marriage characteristics were examined by depression category and nativity specified racial/ethnic groups in order to qualify them as mediators for depression symptoms. Standard errors are shown for all percents in the tables. Logistic regression analyses were used to examine the independent effects of race/ethnicity on various outcomes in mental health-seeking patterns. Adjusted odds ratios (ORs) and 95% confidence intervals (CIs) were computed by using the regression (beta) coefficients and standard errors obtained from the logistic regression models.

Results

Among 8,462 mothers with a child under 1 year old who participated in the ECLS-B home interview, 7,676 (90.7%) had a valid CES-D score and were selected for this study. They represent about 3 million new mothers in 2001. The CES-D scale ranged from 0 to 36, with a median of 4 and a mean of 5.2. It was skewed towards the lower end of the scale. Table 1 shows two sets of proportions: (1) The first two columns describe the sample size and the weighted socio-demographic distribution of all mothers in the study; and (2) the remainder of the table presents distributions of depressive symptomatology categories by socio-demographic variables. The distribution of racial/ethnic groups among the mothers is similar to that of the U.S. population, with 61% non-Hispanic whites, 14.1% of non-Hispanic black, 20.3% of Hispanic, and 3.3% of Asian and Pacific Islander, and 0.6% of Native American. Most of the Hispanic mothers were Mexicans (68%) while the Asian mothers were almost evenly distributed over seven specific ethnic groups. 18.8% of mothers were foreign born and the majority of them lived in urban areas (73.5%).
Table 1

Percentage of CES-D depressive symptom levels (with SE) by socio-demographic characteristics of mothers with child under 1 year old: Early childhood longitudinal study-birth cohort, 9 month interview, 2001

   

Non-depressive

 

Moderate

Severe

 

N

Weighted %

(NDR)

Mild depressive

depressive

depressive

All mothers in ECLS-B

7676

100.0%

59.1%

24.7%

9.7%

6.5%

Mother’s Race/ethnicity

      

 Non-Hispanic White

3918

61.6%

61.5 (1.12)

24 (0.92)

8.8 (0.55)

5.8 (0.55)

 Non-Hispanic Black

1274

14.1%

44.7 (1.52)

29.7 (1.59)

15.3 (1.05)

10.2 (1.03)

 Hispanics

      

 ALL

1246

20.3%

62.2 (1.61)

23.1 (1.46)

8.3 (0.74)

6.5 (0.83)

  Mexican

851

14.2%

61.5 (2.08)

23.6 (1.88)

8.1 (1.02)

6.8 (0.97)

  Puerto Rican

114

1.5%

70.6 (4.77)

17.8 (3.53)

6.2 (1.84)

5.5 (2.73)*

  Cuban

 34

0.6%

64.8 (8.53)

27.1 (7.81)

4.7 (4.61)*

3.4 (3.33)*

  Other Hispanics

247

4%

61.2 (3.35)

22.6 (2.78)

10 (2.04)

6.2 (1.84)

 Non-Hispanic Asian

      

 ALL

918

3.1%

60.0 (1.85)

24.3 (1.57)

11.4 (1.40)

4.3 (0.60)

  Asian Indian

205

0.8%

65.1 (4.22)

20.3 (2.88)

11.9 (3.74)*

2.8 (1.46)*

  Chinese

337

0.7%

64.4 (3.96)

23.4 (3.15)

8.9 (1.8)

3.2 (0.81)

  Filipino

117

0.5%

49 (4.8)

25.3 (3.68)

16.0 (3.8)

9.6 (2.53)

  Japanese

 44

0.2%

72.1 (7.65)

15.3 (5.16)

5.7 (3.35)*

7.0 (4.69)*

  Korean

 66

0.3%

55.6 (8.64)

33.4 (8.34)

8.1 (3.15)*

2.9 (1.69)*

  Vietnamese

 69

0.3%

56.3 (6.46)

21.5 (4.83)

15.6 (5.15)

6.6 (3.23)*

  Other Asian

 80

0.3%

53.1 (5.16)

35.6 (5.57)

10.6 (3.31)

0.6 (0.61)*

 Pacific Islander

 37

0.2%

58.9 (11.03)

29.7 (10.63)

7.2 (3.72)*

4.2 (2.97)*

 Native American

267

0.6%

51.8 (4.31)

28.7 (5.22)

11.8 (2.53)

7.7 (2.00)

Mother’s Nativity

      

 Foreign born

1795

18.8%

63.9 (1.71)

21.9 (1.44)

9 (0.97)

5.2 (0.74)

 US born

5823

81.2%

58 (1.01)

25.3 (0.82)

9.9 (0.49)

6.8 (0.51)

Mother’s Age

      

 <20 years

591

7.7%

46.6 (2.49)

29.8 (2.51)

12.8 (1.84)

10.8 (1.45)

 20–35 years

5972

79%

59.2 (0.88)

24.8 (0.87)

9.8 (0.48)

6.3 (0.43)

 >35 years

1113

13.3%

66 (2.12)

21.1 (1.65)

7.7 (1.01)

5.2 (1.01)

Family annual Income

      

 ≤$20,000

1938

24.5%

45.7 (1.46)

29.7 (1.16)

13.7 (0.97)

11 (0.95)

 $20,001 to $35,000

1752

23%

55.7 (1.74)

25.9 (1.56)

10.4 (1.04)

7.9 (0.95)

 $35,001 to $50,000

1157

15.2%

57.6 (1.87)

27.7 (1.7)

9.9 (1.05)

4.7 (0.71)

 $50,001 to $75,000

1185

15.9%

65.8 (1.8)

22.4 (1.72)

7.9 (1.09)

3.8 (0.82)

 >$75,000

1644

21.5%

74.1 (1.63)

17.1 (1.31)

5.7 (0.82)

3.1 (0.69)

Family socio-economic status**

      

 ses score=1

1364

18%

48.6 (1.63)

28.7 (1.37)

12.7 (1.03)

10.0 (1.11)

 ses score=2

1520

19.5%

48.6 (1.81)

28.8 (1.57)

12.2 (1.14)

10.4 (1.07)

 ses score=3

1521

20.2%

58.1 (1.85)

26 (1.52)

10.5 (1.01)

5.5 (0.71)

 ses score=4

1466

21%

64.8 (1.69)

21.9 (1.47)

8.5 (1.12)

4.9 (0.7)

 ses score=5

1805

21.4%

72.9 (1.39)

19 (1.27)

5.5 (0.71)

2.5 (0.52)

Self reported mother’s health status

      

 Fair/poor

558

6.4%

35.7 (2.83)

28.9 (2.83)

15.7 (1.78)

19.7 (2.56)

 Excellent/very good/good

7117

93.6%

60.7 (0.93)

24.4 (0.8)

9.3 (0.45)

5.6 (0.39)

Has partner at home

      

 Yes

6093

80.4%

63.2 (0.91)

23.2 (0.82)

8.5 (0.45)

5.1 (0.4)

 No

1583

19.6%

42.4 (1.63)

30.8 (1.42)

14.7 (1.17)

12.1 (1.11)

Parity

      

 0

2910

40.2%

59.1 (1.21)

25.6 (1.22)

9.2 (0.66)

6.2 (0.61)

 1+

4742

59.8%

59.2 (1.06)

24 (0.89)

10.1 (0.52)

6.7 (0.55)

Place of Residence

      

 Urban

5589

73.5%

60.2 (1.05)

24.2 (0.93)

9.4 (0.52)

6.2 (0.45)

 Suburban

927

12%

57.1 (1.99)

24.6 (1.43)

12.3 (1.53)

6 (1.06)

 Rural

1160

14.5%

55.5 (1.68)

27 (1.61)

9.1 (0.77)

8.4 (1.01)

*Estimates preceded by an asterisk have a relative standard error of greater than 30% and should be used with caution as they do not meet the standard of reliability or precision.

**Family socio-economic status is a quintile indicator of a composite score consisting of parent’s education, occupation and family poverty levels.

Among all respondents, 59.1% were in the non-depression category, 24.7% had mild depressive symptoms, 9.7% had moderate depressive symptoms, and 6.5% had severe depressive symptoms. When comparing frequency of depressive symptom categories by socio-demographic characteristics, we will hereafter refer to the “frequency of non-depressive category” as the “non-depression rate” (NDR). It is noted that the lower the NDR means the higher the prevalence of some depressive symptoms—i.e., the remaining percentages are distributed among “mild, moderate, or severe” depressive symptoms. US-born mothers in general reported lower NDRs compared to their foreign-born counterparts (NDR: 58.0% vs. 63.9%), which indicates a higher depression rate in the U.S. population. Distributions of depressive symptoms vary significantly among racial/ethnic groups, with non-Hispanic black mothers having the lowest non-depression rate (44.7%), followed by Filipina (49.0%) and Native American (51.8%). Hispanic and Asian mothers in general had a similar prevalence of having any depressive symptoms compared to the general maternal population (NDR: 62.2% for Hispanics and 60.0% for Asian); however, the prevalence varied substantially among the different ethnic groups in Asian mothers.

Filipina mothers reported significantly higher percentage in every depressive symptom category (25.3% in mild depressive, 16.0% in moderate depressive, and 9.6% in severe depressive) compared to the general population of Asian mothers as a whole. Korean and Vietnamese mothers also had elevated rates of depressive symptoms at different levels compared to other Asian mothers (Table 1).

The percentage of depressive symptoms was also elevated in teenage mothers (NDR of 46.6% vs. 59.2% and 66.0% in other age groups); in mothers from low income families (NDR of 45.7% in the mothers from family with income ≤20K vs. 55.7%, 57.6%, 65.8% and 74.1% in higher income groups, respectively); and in mothers with low SES scores (NDR of 48.6% in the two lowest SES score groups compared to 58.1%, 64.8% and 72.9% in the higher SES groups). The presence of depressive symptoms was also associated with mothers’ self-reported physical health: the NDR was 35.7% in mothers reported in fair or poor health compared to 60.7% in mothers reported in good or excellent health. Similarly, mothers who had no partner at home had lower NDRs (42.4% vs. 63.2%). Mothers residing in rural areas had more depressive symptoms (NDR: 55.5%) compared to mothers in urban (NDR: 60.2%) and mothers in suburban areas (NDR: 57.1%). Parity did not make a significant difference in NDR.

Table 2 presents the prevalence of depressive symptom categories by racial/ethnic group stratified by nativity. There are several significant findings when comparing foreign-born with US-born mothers. For non-Hispanic whites, no significant difference was found in the prevalence of depressive symptoms by nativity. Among non-Hispanic Blacks, compared to their US-born counterparts, foreign-born mothers had lower prevalence of depressive symptoms at mild (20.7% vs. 30.6%) and severe categories (4.9% vs. 10.7%). Among Hispanics, foreign-born Hispanic mothers had a lower prevalence in every category of depressive symptoms. Foreign-born Cuban mothers had a lower prevalence of depression symptoms in general, but there were not enough US-born Cuban mothers to compare with in the moderate and severe depression categories. Among Asians and Pacific Islanders, the sample size of foreign-born Asian mothers, except for Filipina mothers, was small and ended up with a wide variance in the frequencies of any depressive symptoms. The overall pattern in Asian mothers was the opposite to that of Hispanic mothers. For most foreign-born Asian mothers (Asian Indian, Chinese, Filipina, Japanese, and Other), their prevalence of any depressive symptom was higher than US-born Asians. Specifically, foreign-born Japanese mothers had significantly higher rates of having some depressive symptoms than those who were US-born (NDR: 51.3% in foreign-born vs. 91.9% in US-born).
Table 2

Percents (with standard errors) of CES-D depression levels by nativity and ethnicity among Asian and Hispanic mothers with a child under 1year old: ECLS-B 2001

 

Foreign born mothers

US born mothers

   

Moderate

   

Moderate

 
 

Non-depressive

Mild depressive

depressive

Severe depressive

Non-depressive

Mild depressive

depressive

Severe depressive

All

63.9 (1.71)

21.9 (1.44)

   9 (0.97)

 5.2 (0.74)

   58 (1.01)

25.3 (0.82)

 9.9 (0.49)

 6.8 (0.51)

Non-Hispanic White

61.5 (4.79)

26.5 (4.57)

8.5 (3.06)*

3.4 (1.96)*

61.5 (1.13)

23.9 (0.9)

8.8 (0.55)

5.8 (0.57)

Non-Hispanic Black

57.6 (4.77)

20.7 (4.7)

16.9 (3.43)

4.9 (2.62)*

43.5 (1.47)

30.6 (1.59)

15.2 (1.09)

10.7 (1.17)

Hispanic

66.3 (2.1)

20.5 (1.83)

7.4 (1.02)

5.8 (1.07)

55.7 (2.81)

27.1 (2.70)

9.56 (1.57)

7.59 (1.27)

 Mexican

66.2 (2.55)

19.9 (2.18)

8.1 (1.33)

5.8 (1.31)

53.9 (3.62)

29.5 (3.33)

8.2 (1.63)

8.4 (1.74)

 Puerto Rican**

74.2 (7.88)

7.4 (3.93)*

6.3 (3.6)*

12.1 (7.38)*

68.9 (6.12)

23 (6.14)

6.1 (2.68)*

2 (1.69)*

 Cuban

60.2 (10.42)

26.7 (8.87)

7.7 (7.33)*

5.5 (5.34)*

72.3 (13.22)

27.7 (13.22)*

 Other Hispanics

66 (3.97)

24.1 (4.00)

5.3 (2.36)*

4.6 (1.94)*

50.8 (5.32)

20.1 (4.64)

19.6 (5.38)

9.5 (4.06)*

Non-Hispanic Asian

58.8 (1.90)

24.0 (1.52)

12.6 (1.59)

4.5 (0.68)

67.66 (6.55)

26.1 (6.44)

3.2 (1.70)*

3.04 (1.83)*

 Asian Indian

64.2 (4.39)

20.6 (2.96)

12.3 (3.86)

2.9 (1.53)*

85 (11.37)

15 (11.37)*

 Chinese

64.1 (4.2)

23.4 (3.25)

9 (1.9)

3.5 (0.89)

67.6 (12.8)

23.3 (12.18)*

9.1 (6.35)*

 Filipino

47.3 (5.54)

25 (4.02)

18.6 (4.19)

9.0 (3.35)*

56 (10.84)

25.1 (9.05)*

7.2 (5.53)*

11.8 (6.78)*

 Japanese

51.3 (11.29)

22.8 (7.66)

11.6 (6.65)*

14.3 (9.02)*

91.9 (5.88)

8.1 (5.88)*

 Korean

59.6 (7.08)

26.1 (6.41)

10.5 (4.28)*

3.8 (2.16)*

42.3 (20.34)

57.7 (20.34)*

 Vietnamese

56.4 (6.65)

20 (4.47)

16.6 (5.44)

7 (3.41)*

56 (25.57)*

44 (25.57)*

 Other Asian

48.9 (5.39)

37.8 (5.67)

12.6 (3.72)

0.7 (0.72)*

76.7 (15.98)

23.3 (15.98)*

Pacific Islander

49.9 (15.15)

35.6 (16.8)*

8 (6.32)*

6.5 (5.66)*

67.5 (12.58)

24 (10.8)*

6.4 (4.66)*

2 (2.1)*

Native American

52.6 (4.99)

28 (6.05)

12.3 (2.76)

7.1 (2.11)

Note. ‒ Quality zero.

*Estimates preceded by an asterisk have a relative standard error of greater than 30% and should be used with caution as they do not meet the standard of reliability or precision

**Foreign born Puerto Rican refers to those who were born in US self-governing Puerto Rico islands.

Table 3 presents the help-seeking patterns by race/ ethnicity in 1,392 (18%) mothers who had moderate to severe depression symptomatology. Because of limited sample size in this sub-population analysis, we did not pursue analysis in specific ethnic groups among Hispanics and Asians. The majority (58.7%) did not feel they needed help and 74.2% did not talk to any health care professional about the symptoms. Minority mothers and foreign-born mothers were about twice as likely not to think they needed help or talk to a doctor compared to non-Hispanic white mothers and US-born mothers, respectively.
Table 3

Percents (with standard errors) and adjusted ORs of mental health-seeking by nativity and selected race/ethnicity among mothers with moderate to severe depression symptomatology: ECLS-B 2001

 

N (mothers with moderate to severe depression

Did not think needed mental health care*

Did not talk to health care professional**

 

symptomatology

%(SE)

AOR***

%(SE)

AOR***

All

1392

58.7 (1.83)

 

74.2 (1.75)

 

Non-Hispanic White

609

48.3 (2.67)

1.0

67.2 (2.9)

1.0

Non-Hispanic Black

359

73.8 (2.39)

2.0 (1.6–2.5)

83.2 (2.75)

2.2 (1.6–3.0)

Hispanic

212

68.3 (4.43)

1.9 (1.5–2.4)

82.3 (3.37)

2.5 (1.8–3.5)

Non-Hispanic Asian

141

76.5 (5.03)

2.2 (1.6–3.1)

87.4 (4.11)

2.5 (1.6–3.8)

Native American

 60

56.8 (6.17)

1.1 (0.9–2.0)

73.0 (6.57)

1.4 (0.9–2.2)

Foreign Born

287

77.1 (3.85)

1.9 (1.5–2.4)

88.3 (2.92)

2.4 (1.7–3.4)

US Born

1094

55 (1.87)

1.0

71.4 (1.86)

1.0

*During the past 12 months, have you felt, or has anyone suggested, that you needed help for any emotional or psychological problem?

**In the past months, have you talked with a psychiatrist, psychologist, doctor or counselor for any emotional or psychological problem?

***Adjusted Odds Ratio adjusted for family income levels.

Table 4 presents the distribution of the potential socio-demographic risk factors by racial/ ethnic groups and nativity for all ECLS-B mothers. It shows that US-born mothers had higher percentages in lacking a partner at home and living in rural areas compared to foreign-born mothers. Their SES status was better than foreign-born mothers (low SES: 33.9% in US-born vs. 52.9% in foreign-born). Among specific racial/ethnic groups, results indicated there were very limited foreign-born non-Hispanic white mothers in any of the high risk groups. Among non-Hispanic Blacks, foreign-born mothers had lower rates in the low SES and lack of partner at home groups compared to their US counterparts. Among Hispanics, Mexican foreign-born mothers were less likely to be teenagers (7.4% vs. 19.2%), or lack of partner at home (12.0% vs. 26.8%) compared to US-born Mexican mothers; however, these foreign-born mothers were more likely to be poor (77.5% vs. 55.6%). Among Asians, the overall prevalence of any socio-demographic risk factors was low. Foreign-born Asian mothers had even lower frequencies in many of the socio-demographic factors compared to their US-born counterparts. They tended to be less likely to be teenage mothers (2.6% vs. 11.0%), and less likely to lack a partner at home (4.9% vs. 12.6%), but slightly more likely to have a physical health complaint.
Table 4

Percent (with SE) of new mothers with risk factors for depression by race/ethnicity and nativity: ECLS-B 2001

 

Foreign-born mothers

US Born Mothers

 

Teenage Mother

Live in rural area

Lower SESa

No partner at home

Poor physical health

Teenage Mother

Live in rural area

Lower SESa

No partner at home

Poor physical health

All

6.5 (0.83)

2.4 (0.46)

52.9 (1.89)

12.4 (1.26)

8.1 (0.98)

8 (0.51)

17.2 (1.17)

33.9 (1.21)

21.2 (0.95)

6 (0.41)

Non-Hispanic White

1.5 (1.13)

6.5 (2.41)

16.3 (3.78)

4.2 (2.22)

1.6 (1.01)

5.3 (0.58)

21 (1.37)

26 (1.46)

11.5 (0.72)

4.9 (0.5)

Non-Hispanic Black

5.3 (2.74)

35.6 (5.94)

29.7 (4.53)

5.6 (2.63)

15.1 (1.38)

8.3 (1.34)

59.8 (1.68)

63 (1.7)

8.1 (1.04)

Hispanic

          

 All

8.6 (1.11)

1.9 (0.49)

69.8 (2.30)

13.8 (1.47)

10.5 (1.41)

15.9 (1.84)

4.3 (1.16)

50.3 (2.85)

26.8 (2.34)

10.1 (1.54)

 Mexican

7.4 (1.08)

2.7 (0.69)

77.5 (2.44)

12 (1.69)

12.2 (1.77)

19.2 (2.35)

5.1 (1.4)

55.6 (3.12)

26.8 (2.87)

11.5 (1.98)

 Puerto Rican*

13.7 (7.61)

33.5 (12.68)

19.6 (8.81)

15.9 (6.44)

7.3 (3.42)

0.2 (0.16)

51 (8.74)

30.1 (6.04)

6.2 (2.68)

 Cuban

4.4 (4.1)

22.2 (11.27)

22 (10.11)

6.4 (4.26)

8.8 (8.13)

15.4 (9.96)

 Other Hispanics

12 (2.96)

58.4 (5.02)

17.1 (3.46)

5.2 (2.21)

11.1 (3.38)

4.9 (3.82)

35.4 (7.87)

27.4 (5.96)

6.5 (3.37)

Non-Hispanic Asian

          

 All

2.6 (1.29)

2.0 (0.58)

16.3 (2.06)

4.9 (1.14)

4.6 (0.85)

11.0 (4.21)

3.1 (1.90)

17.0 (4.36)

12.6 (4.18)

2.7 (1.60)

 Asian Indian

4 (3.89)

3.2 (1.26)

14.5 (5.51)

1.6 (1.01)

1.4 (0.7)

8.9 (9.32)

15.6 (11.79)

8.9 (9.32)

 Chinese

0.5 (0.47)

1.0 (0.63)

8.4 (2.21)

1.5 (0.86)

2.9 (0.94)

5.3 (5.22)

9.1 (6.35)

9.1 (6.35)

 Filipino

2.6 (2.04)

1.9 (1.11)

16.2 (4.06)

12.2 (4.2)

6 (3.01)

4.9 (3.41)

2.1 (2.04)

20.6 (9.06)

7.8 (5.96)

 Japanese

2.9 (2.82)

3 (2.84)

11.8 (8.04)

3.9 (3.87)

3.9 (3.87)

 Korean

2.4 (1.62)

2.8 (2.76)

1.1 (1.08)

5.2 (3.36)

4.8 (3.29)

10.2 (10.22)

19.5(13.83)

27.1 (16.1)

7.6 (7.77)

 Vietnamese

2.7 (2.76)

39.4 (7.37)

9.5 (4.19)

12.6 (4.86)

 Other Asian

5 (2.66)

1.6 (1.2)

37.5 (6.27)

10.8 (4.32)

9.2 (3.28)

66 (12.47)

51.6(16.23)

35 (21.39)

Pacific Islander

67 (12.25)

35.4 (17.75)

15.7 (9.89)

5.2 (6.54)

26.2 (12.98)

12.9 (10.7)

2 (2.1)

Native American

10 (2.01)

25 (4.02)

48.4 (5.24)

23.5 (3.87)

17.2 (3.83)

Note. ‒ Quality zero.

*Foreign born Puerto Rican refers to those who were born in US self-governing Puerto Rico islands.

Discussion

In this report, we estimate the prevalence of maternal depressive sympotomatology and its association with socio-demographic and economic characteristics among the major racial/ethnic groups by nativity in the U.S. We also assess mental help-seeking patterns in a sub-group of mothers who had moderate or severe depressive symptoms. Our descriptive tabulations demonstrate that the prevalence of depressive symptoms is higher in US-born mothers compared to foreign-born in the White, Black and Hispanic populations. Foreign-born Asian mothers, however, have a higher prevalence of depressive symptoms than their US-born counterparts, despite the lower frequency of socio-demographic risk factors. The prevalence also varies widely within different ethnic groups of the Asian/Pacific Islander population. Our findings also suggest that mental health services utilization remains a major problem among minorities and immigrant populations.

Our results also confirm that non-Hispanic black mothers have the highest prevalence of any depressive symptomatology, and it is the US-born mothers who bore most of the weight. Risk factor analysis indicates that US-born non-Hispanic Black mothers, compared to their foreign-born counterparts and mothers from other racial/ethnic groups, are more likely to have no partner at home and have low SES status. These findings are consistent with previous studies on physical and mental health status in non-Hispanic blacks by nativity [3840].

Similar to the foreign-born black mothers, we found Hispanic immigrant mothers had lower prevalence of depression symptomatology compared to their US-born counterparts, despite the fact that they are more likely to be poor. The unfavorable condition could be offset by their social support: they are less likely to be teenage mothers and more likely to have a partner at home. These findings are in congruence with the risk and protective factors studies in Latina women during pregnancy and this report extend this finding to the post-partum period [41, 42].

Our study demonstrates that the prevalence rates among Asian mothers vary considerably, ranging from low (Chinese and Asian Indian) to very high (Filipina). These are consistent with a growing literature indicating that the Asian/Pacific Islander population in the U.S. is a highly heterogeneous group in terms of history, culture and SES [43]. Our results also show that, contrary to findings in other racial groups, Asian immigrant mothers are more likely to report depressive symptoms compared to Asian American mothers. This phenomenon is most obvious in Japanese mothers, but also in Asian Indian, Filipina, and other Asian groups. Unfortunately we don’t have enough sample size to test for statistically significant differences between US-born and Foreign- Born Asian mothers in sub-ethnic groups except for Japanese mothers. These intriguing patterns of depression symptoms cannot be fully explained by the social-demographic risk factors. For example, although Filipina mothers fare much better in terms of SES compared to non-Hispanic black mothers, they had similar depression rates. Meanwhile, a previous study has shown the high prevalence of low birth weight and infant mortality in Filipino immigrant populations in US [40]. Still, it is unclear why Filipina mothers bear extra weight in depression symptoms among Asians.

Our results on the help-seeking patterns in mothers with moderate to severe depressive symptomatology show that in general only one fourth of mothers who are at risk have talked to health care professionals. This is consistent with WHO’s estimate that fewer than 25% of those affected by depression have access to effective treatment [1]. The minorities and foreign-born are at higher risk for having limited access to mental health services utilization. A number of explanations have been offered in previous studies to account for the higher risk of low mental health services utilization by the minorities and the foreign born. Among them there are theories that the social stigma of a mental health problem is more severe in minority populations [44, 45], and lower acculturation levels are associated with greater stigma [46, 47]. Unfortunately there are no measures of stigma in the ECLSC-B data, thus interpretations would be beyond limitations of the data in this project.

One of the benefits of having an additional question on the “need” of mental health care is to detect the discrepancies between the frequency of need and that of actual utilization. We found that a higher percentage of mothers with moderate or severe depressive symptoms claimed they needed the services (41.3%) than the percentage who talked to health care professionals (25.8%). In addition, compared to non-Hispanic White and US-born mothers, minority and foreign-born were 2.2–2.5 times less likely to lack of mental health access, while their likelihood of thinking they need mental health services compared to non-Hispanic White or US born mothers was relatively lower at 1.9 to 2.2 times across racial groups. In other words, the gap was bigger for minorities and foreign-born compared to non-Hispanic White mothers than the gap in need for mental health utilization compared to the gap for needing mental health care.

This study has several limitations. Most depressed mothers probably were less enthusiastic about the home interview, which may have led to an overall selection bias in participation. This may explain why the prevalence of depression symptoms by racial/ethnic group is lower in our study than other studies of perinatal depression in general. Another possible selection bias is that the self-administrated questionnaire was only available in English and Spanish. Mothers who cannot read English were excluded from this study. We also cannot include mothers with invalided CESD scores and those whose infant died before the 9 month interview. The timing of the 9-month interview limited us to study mothers in the 6th to 12th months after childbirth, when the most vulnerable time for post-partum depression (<3 month after delivery) has passed for most mothers. Therefore we could have under-estimated the prevalence of post-partum depression. The CES-D has been used in studies that involve multiple racial/ethnic populations and both sexes as a useful screening tool for depression [48, 49]. However, it has not been validated for post-partum depression. Finally, even though the ECLS-B is the first national survey focusing on new mothers and has over-sampled certain Asian ethnic groups, there are still not sufficient data to conduct hypothesis testing analysis on how multiple risk behavior affect depression prevalence within each racial/ethnic/nativity group. Due to the limited scope of the survey, we also cannot distinguish foreign-born mothers by their immigration status and length of stay in the US.

This study, using a national database to estimate the prevalence and the socio-demographic risk factors for post-partum depression, shows significant differences for depressive symptomatology in new mothers by race/ethnicity and nativity. Our manuscript is the first social-epidemiology study to report the burden of maternal depression in 30 sub-ethnic/nativity groups using recent national survey data. Previous epidemiological research has found that Asian Americans have a lower prevalence rate of depression than other racial groups [40, 50]; however, our research suggests that this “global estimate” masks sub-ethnic group differences. We think this is an important first step to understanding the scale of the problem. Future research is needed to delineate the determinants of perinatal depression in certain high prevalence sub-ethnic and nativity groups. More efforts are needed to overcome the barriers in mental health services access and utilizations, especially in minority and foreign-born populations.

Acknowledgments

The authors wish to thank Drs. Michael Kogan and Gopal Singh, in the Office of Data and Program Development, Maternal and Child Health Bureau (MCHB) of HRSA for their comments on the earlier draft of this article. We are grateful for the IPA funding for Dr. Huang to work on the manuscript from the Office of Data and Program Development, MCHB/HRSA. And the access to ECLS-B 9 month data provided by the MCHB and Department of Education.

Copyright information

© Springer Science+Business Media, LLC 2006