Research Paper

Archives of Psychiatric Nursing 30 (2016) 7–12

Contents lists available at ScienceDirect

Archives of Psychiatric Nursing

journal homepage: www.elsevier.com/locate/apnu

Racial and Ethnic Differences in Prenatal Life Stress and Postpartum

Depression Symptoms

Cindy H. Liu a,⁎, Rebecca Giallo b, Stacey N. Doan c, Larry J. Seidman a, Ed Tronick d a Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA b Murdoch Children’s Research Institute, Royal Children’s Hospital, Parkville, VIC, Australia c Claremont McKenna College, Claremont, CA d University of Massachusetts Boston, Children Development Unit, Boston, MA

a b s t r a c t

This study determined the risk of core depression symptoms based on life stress domains during pregnancy and whether stressors varied by race/ethnicity. The sample consisted of 2,344 White, African American, Hispanic, and Asian/Pacific Islander (API) Massachusetts women who recently gave birth. African Americans and Hispanics who endorsed high relational and high financial stress were more likely to report high depressed mood and loss of interest; high physical stress was associated with high depressed mood among API. Screening based on life stress domains may be informative in determining risk for core depression symptoms during the postpartum period especially for minority groups.

© 2015 Elsevier Inc. All rights reserved.

Postpartum depression (PPD) affects approximately 20% of women worldwide (Gavin et al., 2005), and is marked with core symptoms in- volving depressed mood and/or anhedonia (loss of interest in previous- ly enjoyed activities) (Sibitz et al., 2010a). Although PPD affects women from all backgrounds, prevalence rates among U.S. race/ethnic groups vary, with mixed findings. Studies have shown African American and Hispanic women tend to have higher (Howell, Mora, Horowitz, & Leventhal, 2005), lower (Wei, Greaver, Marson, Herndon, & Rogers, 2008), or similar (Yonkers et al., 2001) PPD rates compared to Whites (Liu & Tronick, 2013a). Recent evidence indicates a higher likelihood of PPD among Asian Americans (Gavin et al., 2011; Liu & Tronick, 2013a). Socioeconomic status (SES) may be a reason for these discrep- ancies (Beeghly et al., 2003; Hobfoll, Ritter, Lavin, Hulsizer, & Cameron, 1995; Liu & Tronick, 2013a; Rich‐Edwards et al., 2006). Because minor- ity groups often experience greater social and economic adversity, un- derstanding the life experiences that accompany these circumstances is important (Thoits, 2010).

Focusing on stressful life events during pregnancy associated with PPD (e.g., moving, a family death, poverty) may help clarify the mecha- nisms underlying differential rates of PPD among U.S. racial and ethnic minority groups, and the relationship between prenatal life stress and postpartum depression (Da Costa, Larouche, Dritsa, & Brender, 2000; Lancaster et al., 2010; O’Hara & Swain, 1996; Robertson, Grace, Wallington, & Stewart, 2004). For example, African Americans (AAs) and Hispanics show higher prevalence of perinatal stress experiences and depression than Whites, with one study reporting that 51% of AA and Hispanic women (n=148) had elevated depression symptoms

⁎ Corresponding Author: Cindy H. Liu, Ph.D., Beth Israel Deaconess Medical Center, Har- vard Medical School, 75 Fenwood Road, Boston, MA 02115 USA.

E-mail address: [email protected] (C.H. Liu).

http://dx.doi.org/10.1016/j.apnu.2015.11.002 0883-9417/© 2015 Elsevier Inc. All rights reserved.

and incidence of negative life events (Zayas, Cunningham, McKee, & Jankowski, 2002). In another study, a greater number of stressors were associated with postpartum depressed mood for Whites, AAs, His- panics, and Asian/Pacific Islander (APIs) and postpartum anhedonia only for Whites and Hispanics (Liu & Tronick, 2013a, 2013b, 2013c) while an earlier study on AAs and Whites found no association between prenatal life stress and PPD (Ritter, Hobfoll, Lavin, Cameron, & Hulsizer, 2000). Notably, life events may be highly contextualized by cultural his- tories and experiences, beliefs, and practices, leading to their differential impacts (Hobfoll, 1998).

This study examines the prevalence of different domains of life stress experienced by pregnant racial/ethnic minority women and its associa- tion with subsequent postpartum mood symptom, specifically de- pressed mood and anhedonia, in a population of Massachusetts women (Liu & Tronick, 2013b). We consider life stress based on domain (i.e., financial, relational, physical health), rather than on individual stressors or the total number of stressors, as life domains may be more predictive of later PPD (Liu & Tronick, 2013b). If individuals within dif- ferent cultural communities are sensitive to different domains, organiz- ing stress experiences by context may be useful for understanding how life stress confers risks to specific groups and areas of heightened PPD risk among minority women.

METHODS

Participants

Data were drawn from the 2007–2008 PRAMS (Pregnancy Risk As- sessment Monitoring System), a population-based survey administered to postpartum women in Massachusetts. The PRAMS is a collaborative

8 C.H. Liu et al. / Archives of Psychiatric Nursing 30 (2016) 7–12

surveillance project conducted by the Centers for Disease Control and Prevention (CDC) and state health departments to monitor maternal behaviors and experiences of women before, during and after live birth pregnancies. The Massachusetts Department of Public Health (MDPH) provided the dataset for this study. Institutional Review Board (IRB) approvals from the relevant institutions were obtained to conduct the data collection and analysis.

Those who participated in this population-based survey were ran- domly sampled women from Massachusetts who had recently had a live birth and were provided birth certificates. Women who were 2–6 months postpartum were selected to receive up to three mailed paper surveys; those who did not respond to these surveys were contacted by telephone. Response rates for 2007 and 2008 were 71%, over the CDC-recommended response rate of 65% for public data sharing. In order to represent the Massachusetts birth population from 2007–2008, this data was weighted based on maternal demographics, which accounted for non-response and adjusted for sampling probabilities.

Measures

Demographic Information The birth certificate provided maternal nativity, maternal age, edu-

cation, and race/ethnicity variables. Based on self-report, women were classified as Hispanic or non-Hispanic. Non-Hispanic women were cat- egorized in one of the following groups: White, AA, API, and American Indian/Alaskan Native. Mean infant age at survey completion was 9.3 weeks; groups showed no significant differences in infant age. House- hold income was based on participant report (“total household income before taxes in the 12 months before the new baby was born”).

PrenataL Life Stressors Through the PRAMS survey, respondents indicated “yes” or “no” to

events that occurred during the last 12 months before the new baby was born (e.g., “I moved to a new address,” “I had a lot of bills to pay,” “I got separated or divorced from my husband or partner.”) (Table 1). Prenatal life stress domains were the main predictors of depression and anhedonia in this study. The calculation of these stress domains were based upon previously derived principal components analyses (PCA) conducted on a similar, diverse, population-based sample of women (Liu & Tronick, 2013b). In the previous study, 2 stressors were dropped based on low endorsements (“homelessness, husband or

Table 1 PRAMS prenatal stress survey items.

Stress (Specific Events) Items

Relational Separation or Divorce “I got separated or divorced from my husband or

partner.” Arguments at Home “I argued with my husband or partner more than

usual.” Husband Does Not Want Pregnancy

“My husband or partner said he didn’t want me to be pregnant.”

Physical Fighting “I was in a physical fight.” Drug Problems of Close Friends or Family

“Someone very close to me had a bad problem with drinking or drugs.”

Financial Moved “I moved to a new address.” Husband Lost Job “My husband or partner lost his job.” Mother Lost Job “I lost my job even though I wanted to go on

working.” Trouble Paying Bills “I had a lot of bills I couldn’t pay.” Physical Family Illness “A close family member was very sick and had to

go into the hospital.” Death of Close Friends or Family

“Someone very close to me died.”

Note: Two additional items pertaining to homelessness and jail were dropped from the analyses because of the low numbers endorsed.

partner in jail”). The current sample also had too few of these two items for inclusion. As such, the PCA on 11 stressors from the PRAMS survey produced three domains: Relational, Financial, and Physical Health (Table 1). We used a previously published approach in calculat- ing the domains: a sum of the items endorsed within each domain was calculated and high and low-risk stresses within each domain were de- termined based on the following cut-off scores: (1) Low Relational Risk b2; High Relational Risk ≥2; (2) Low Financial Risk b2; High Financial Risk ≥2; and (3) Low Physical Health Risk b1; High Physical Health Risk ≥1 (Liu & Tronick, 2013b).

Depression Symptoms Respondents were also asked to indicate “always,” “often,” “some-

times,” “rarely,” or “never,” to an item on depressed mood (“Since your new baby was born, how often have you felt down, depressed, or hopeless?”) and an item on loss of interest (“Since your new baby was born, how often have you had little interest or little pleasure in doing things?”). Although a single item is used to assess each of the two symp- toms, these items are extracted from the Patient Health Questionnaire, a commonly used and validated depression screener in primary care set- tings (Arroll et al., 2010; Kroenke, Spitzer, & Williams, 2003).

In following previous published work, endorsements of either “al- ways” or “often” were recoded as having “high depressed mood” or “high loss of interest;” responses of sometimes, rarely, or never were recoded as those with “low depressed mood” or “low loss of interest” (Liu & Tronick, 2013a, 2013b, 2013c).

Statistical Analyses Variables with less than a 100% response rate included maternal ed-

ucation (99.7%), household income (90.0%), stress (98.2%), depressed mood (65.9%), and loss of interest (65.6%) resulting in an unweighted study sample of 2,344 individuals. These rates are consistent with previ- ously published analyses on depressed mood and loss of interest. Non- responders tended to be non-U.S. born, less educated, younger, ethnic minorities, with lower household income. While the difference between responders and non-responders may be a study limitation, these vari- ables typically indicate higher PPD risk; therefore, any association be- tween these factors and PPD from the findings may be a more conservative estimate than if all had responded. To maintain this con- servative approach, we used listwise deletion to handle the missing data as we have done in previous published work in this area (Liu & Tronick, 2013a, 2013b, 2013c).

In identifying the specific role of life stress during pregnancy, we controlled for maternal age, household income, maternal education, maternal nativity, pregnancy intention, gestational diabetes, and infant NICU stay, all which have showed an association with PPD (Beck, 2003; Johnstone, Boyce, Hickey, Morris-Yatees, & Harris, 2001; O’Hara & Swain, 1996; Robertson et al., 2004).

To account for the stratified and weighted sample, the data was an- alyzed using the Complex Samples module of SPSS version 22.0 (SPSS Inc., Chicago, IL). Prevalence estimates within each group were generat- ed according to stressors as predictors. A series of logistic regression analyses were performed, all of which adjusted for socioeconomic sta- tus (maternal age, maternal education, household income) and other potential covariates (pregnancy intention, gestational diabetes, infant NICU stay) with stress domains as predictors. We conducted overall and race-stratified models to obtain the likelihood of endorsing/ predicting high depressed mood and high loss of interest.

RESULTS

Descriptive Characteristics

Table 2 presents sociodemographic data regarding this sample of 2,344 participants. In general, on average, Whites and APIs were older, had more years of education, and had higher levels of income relative

Table 2 Weighted percentage of mothers who completed the MA PRAMS from 2007–2008, by sociodemographic factors according to race/ethnicity.

Total White Black Islander

Hispanic Asian/Pacific Islander

(n=2,344) (n=799) (n=464) (n=596) (n=516)

Maternal Age b20 6.0 4.7 9.1 14.0 2.6 20–29 39.5 36.0 49.2 56.5 35.0 30–39 50.8 55.2 38.6 27.1 59.0 40+ 3.8 4.1 3.1 2.4 3.5 Maternal Education bHigh School 9.0 6.0 9.3 27.1 7.1 High School 23.8 21.0 37.7 36.9 15.4 Some College 18.5 17.4 29.9 21.5 13.2 College Graduate 48.7 55.6 23.1 14.5 64.2 Income b10,000 12.7 7.2 26.1 37.7 10.8 10,000–14,999 6.4 4.8 10.6 13.0 6.6 15,000, 19,999 3.6 2.5 6.5 7.3 4.7 20,000–24,999 4.9 3.5 11.0 9.4 5.2 25,000–34,999 9.0 8.4 14.8 10.3 7.1 35,000–49,999 7.9 7.3 11.7 7.3 9.9 N50,000 55.6 66.3 19.3 14.9 55.8 Maternal Nativity U.S. Born 74.6 89.9 54.2 35.2 16.5 Non-U.S. Born 25.4 10.1 45.8 64.8 83.5

Table 3 Weighted percentage of mothers who completed the MA PRAMS from 2007-2008, by stressful events and depression symptoms according to race/ethnicity.

Total White African American Islander

Hispanic Asian/ Pacific Islander

(n=2,344) (n=799) (n=464) (n=596) (n=516)

No. of Stressful Events 0 33.8 34.9 21.7 26.0 47.1 1–2 42.6 43.2 39.6 41.2 41.5 3–5 19.6 18.3 29.3 27.3 9.9 6–13 4.1 3.5 9.4 5.5 1.5 Specific Stressful Events Family illness 22.4 24.4 20.7 19.1 11.4 Separation of divorce 5.9 4.0 15.7 12.5 3.3 Moved 32.9 31.5 34.7 39.8 32.1 Husband lost job 9.1 8.3 14.9 12.3 6.1 Mother lost job 7.5 5.8 13.3 14.5 6.3 Arguments at home 22.1 19.3 36.2 32.0 17.0 Husband does not want pregnancy

7.3 5.9 16.9 10.9 4.9

Trouble paying bills 18.9 16.9 36.9 26.5 8.3 Physical fighting 2.1 1.2 6.1 5.3 1.1 Drug problems of close friends or family

11.7 13.3 9.9 9.4 2.4

Death of close friends or family

15.5 16.3 19.1 14.5 5.1

Relational Stress1

Low 96.0 97.1 89.3 92.1 98.8 High 4.0 2.9 10.7 7.9 1.2 Financial Stress1

Low 83.9 85.8 72.9 75.6 90.3 High 16.1 14.2 27.1 24.4 9.7 Physical Stress1

Low 71.4 69.6 69.2 73.5 86.9 High 28.6 30.4 30.8 26.5 13.1 Depression Items2

Feelings of Depression

Low 92.4 93.5 89.1 87.7 92.4 High 7.6 6.5 10.9 12.3 7.6

Loss of Interest Low 91.2 93.6 85.2 86.6 81.8 High 8.8 6.4 14.8 13.4 18.2

1 To determine high and low risk stresses within each domain, use followed the fol- lowing cut-off scores based on number of endorsed stressors within each domain from 12 months prior to the birth of the baby: (1) Low Relational Risk b2; High Relational Risk ≥2; (2) Low Financial Risk b2; High Financial Risk ≥2; and (3) Low Physical Health Risk b1; High Physical Health Risk ≥1.

2 Endorsements of either always or often were recoded as having “high depressed mood” or “high loss of interest.” Responses of sometimes, rarely, or never were recoded as those with “low depressed mood” or “low loss of interest.”

9C.H. Liu et al. / Archives of Psychiatric Nursing 30 (2016) 7–12

to AAs and Hispanics. The majority of APIs (83.5%) and Hispanics (64.8%) were non-U.S. born compared to White (10.1%) and AAs (45.8%).

The majority of individuals in each group reported 0–2 stressors (Table 3). A greater proportion of AAs and Hispanics compared to Whites and APIs reported 3+ stressors. Overall, moving during preg- nancy was the most common stressor, with approximately one third of respondents endorsing this stressor. The second and third most com- mon stressors were a family member who was ill and arguments at home. For AAs and Hispanics, the top three stressors were moving, ar- guments at home and trouble paying bills. With respect to domains, physical health stress was the most prevalent stress domain, followed by financial and relational stress, for all groups. AAs and Hispanics showed comparable rates of high physical health and high financial stress. Compared to Whites, AAs, and Hispanics, APIs had the lowest rates of high physical health, high financial, and high relational stress. Indeed, APIs were half as likely to endorse high physical health stress compared to all other groups.

Associations Between Stressful Life Events, Depressed Mood, and Anhedonia

Compared to Whites and APIs, AAs and Hispanics had less education and income. Models in Table 4 show the association between domains and depressed mood; no differences were observed after adjusting for covariates. High relational and high financial stress significantly predict- ed high depressed mood overall, and race-stratified analyses showed that only AAs and Hispanics with high relational and high financial stress were more likely to report high depressed mood. While the asso- ciation between high relational stress and depressed mood was statisti- cally significant for APIs, the wide confidence interval suggests power that may be too low to interpret. High physical stress was associated with high depressed mood among API.

Table 5 presents additional models examining the association be- tween stress domains and loss of interest overall and for each group be- fore and after adjusting for covariates. In general, stress domains did not predict loss of interest among Whites. However, high relational stress was associated with AAs’ loss of interest, and high financial stress was associated with Hispanics’ and APIs’ loss of interest.

DISCUSSION

This study explored the associations between prenatal life stress on postpartum depressed mood and anhedonia in different racial and eth- nic groups in order to shed light on areas of prevention in postpartum mood disorders. Prenatal life stress was common to all groups, but in varying degrees, suggesting that women from all groups experience some sort of adversity during their pregnancy. Specifically, moving and a family member’s death were most prevalent across all groups. While arguments at home and trouble paying bills – relational and fi- nancial concerns – were the next set of common stressors, they were particularly more prevalent among AAs and Hispanics. These rates were slightly lower than previous PRAMS findings in 2000 (Lu & Chen, 2004). However, our study was similar in that AAs endorsed stressors at a higher rate compared to other groups. The relative prevalence of these stressors among AAs and Hispanics is concerning since the nega- tive and chronic nature of these stressors may be associated with worse outcomes in these groups (Myers, 2009; Williams, Yu, Jackson, & Anderson, 1997).

Table 4 Race/ethnicity stratified logistic regression showing adjusted odds of high depressed mood per predictor by race/ethnic group.

Overall White African American Hispanic Asian/Pacific Islander

Stress Domain OR CI OR CI OR CI OR CI OR CI

High Relational Unadjusted 2.3* 1.2–4.6 1.8 0.4–6.9 4.2** 1.9–9.5 3.4** 1.6–7.4 11.5* 1.1–124.12

Adjusted1 2.4* 1.2–4.7 1.6 0.5–6.0 3.7** 1.7–7.8 2.8** 1.3–6.0 8.0* 1.1–55.82

High Financial Unadjusted 2.5*** 1.6–4.0 2.3* 1.0–5.2 4.0*** 1.9–8.1 2.6** 1.5–4.7 1.5 0.5–4.6 Adjusted1 2.5*** 1.6–4.0 2.3* 1.1–5.1 3.3**** 1.7–6.4 2.8*** 1.6–4.8 1.5 0.5–4.3

High Physical Unadjusted 1.3 0.8–2.1 1.1 0.5–2.2 1.3 0.7–2.8 1.2 0.7–2.2 3.6** 1.6–8.1 Adjusted1 1.3 0.8–2.1 1.1 0.6–2.3 1.4 0.7–2.8 1.3 0.7–2.3 3.0** 1.4–6.6

†pb0.1, *pb.05, **pb.01, ***pb.00. 1 Covariates included maternal age, household income, maternal education, maternal nativity, pregnancy intention, gestational diabetes, and infant NICU stay. 2 Wide confidence internal values likely due to small cell size among Asian/Pacific Islanders and should be interpreted with caution.

10 C.H. Liu et al. / Archives of Psychiatric Nursing 30 (2016) 7–12

Although high relational and high financial stressors were associated with depressed mood, this result was specific to AAs and Hispanics, not Whites. Thus, stress in particular domains may be particularly burden- some to AAs and Hispanics, and sociodemographic status does not fully account for these differences.

That high relational and high financial stress from pregnancy predicts depressed mood and loss of interest (using high relational for AA and high financial for Hispanics) in AAs and Hispanics, but not Whites, is noteworthy. Why might relational and financial concerns confer greater risk for these symptoms among AAs and Hispanics? One interpretation may stem from the importance of relationships and networks for AAs and Hispanics. Evidence suggests that the impact of low social support on the risk of major depression is more pronounced among those with larger social networks and/or who tend to rely on emotional support (Kendler, Myers, & Prescott, 2005). AAs may be more likely to cope with economic and interpersonal problems with their family through in- formal help (Broman, 1996; Zayas et al., 2002). Similarly, family support and income were the two best predictors of depression in a sample of Mexican American women (Vega, Kolody, Valle, & Weir, 1991). Although the relational issues measured in our study largely relate to the immediate family, the impacts of these social disruptions may speak to the importance of relationships in these groups. Alternatively, among AAs or Hispanics, the circumstances around the specific relational and financial concerns may be more serious or have a greater adverse effect on other aspects of life, com- pared to Whites. For example, the reasons and consequences involved in “moving to a new address,” could be more problematic and stressful for AAs and Hispanics regardless of socioeconomic status.

High physical stress was associated with depressed mood mainly for APIs, but not with loss of interest. APIs may have a very different inter- pretation of stress compared to other groups, with evidence indicating greater family involvement among Asians Americans in caring for a family member’s physical illness (Nilchaikovit, Hill, & Holland, 1993). Given the tendency for Asians to somaticize stressful experiences,

Table 5 Race/ethnicity stratified logistic regression showing adjusted odds of high loss of interest per p

Overall White Af

Stress Domain OR CI OR CI O

High Relational Unadjusted 1.9† 0.9–4.0 2.7 0.8–9.2 2. Adjusted1 2.0† 0.9–4.1 2.2 0.6–8.3 2.

High Financial Unadjusted 1.8** 1.2–2.7 1.8 0.8–3.8 1. Adjusted1 1.8** 1.2–2.7 2.0† 1.0–4.0 1.

High Physical Unadjusted 0.9 0.6–1.4 0.9 0.4–1.8 1. Adjusted1 0.9 0.6–1.4 0.9 0.5–1.9 1.

1 Covariates included maternal age, household income, maternal education, maternal nativi † pb0.1, *pb.05, **pb.01, ***pb.001.

especially those pertaining to mental health (Bauer, Chen, & Alegría, 2012), their endorsements of illness and health related problems may also simultaneously encompass and mask other related life stresses.

All of these associations take into account key sociodemographic characteristics. Specifically, high relational and high financial stress may serve as risks for core depression symptoms, regardless of reported income, education, marital status, and age. It should also be noted that what is considered “high” stress in each of the domains is based upon a relatively conservative domain specific cut-off score. For instance, de- pending on the domain, “high” stress indicates an endorsement of ei- ther 1 or 2 stressors. Given these measures, our findings suggest that even minimal levels of prenatal stress may impact depressed mood for certain groups.

The lack of associations between stress domains and depression symptoms among Whites was surprising. Previous findings from this data set suggest that higher numbers of stressors are associated with depressed mood in Whites (Liu & Tronick, 2013a, 2013b, 2013c). De- spite the large sample size, we were unable to compare the predictive strength based on the sum of stressors and the domain of stressors while stratifying by race/ethnicity, as it would have significantly re- duced statistical power. Previous work has sought to compare risk ap- proaches in determining the best predictor of outcomes (Burchinal, Roberts, Hooper, & Zeisel, 2000); however, there has been very little in- terpretation on how the experience of such risks play a role in the out- come. One possibility is that Whites perceive their experience of stress differently from minorities, and that the amount rather than the type of stress experienced has a greater impact on depressed mood. In gener- al, AAs and to a lesser extent, other minorities, report more stressful events across different domains compared to Whites (Hatch & Dohrenwend, 2007); perhaps a higher number of stressors, regardless of their particular domains, may be particularly salient to White women.

We chose to examine the specific depressive symptoms of anhedo- nia given the paucity of surrounding research, even though it is a core

redictor by race/ethnic group.

rican American Hispanic Asian/Pacific Islander

R CI OR CI OR CI

6* 1.2–5.8 1.6 0.7–3.8 1.4 0.2–12.9 3* 1.1–5.0 1.5 0.6–3.3 1.4 0.6–3.3

2 0.6–2.2 2.5** 1.4–4.3 1.0 0.4–2.3 2 0.7–2.3 2.4** 1.4–4.0 2.4** 1.4–4.0

1 0.6–2.0 0.7 0.4–1.3 0.9 0.4–2.1 2 0.7–2.2 0.8 0.4–1.4 0.8 0.4–1.4

ty, pregnancy intention, gestational diabetes, and infant NICU stay.

11C.H. Liu et al. / Archives of Psychiatric Nursing 30 (2016) 7–12

symptom of major depression and other major mental illness. Our cur- rent analyses do show some associations between prenatal life stress and anhedonia. Given previous findings on this dataset showing greater endorsement of anhedonia among non-US born individuals and greater prenatal life stress as predictive of anhedonia only for White and His- panics (Liu & Tronick, 2013a, 2013b, 2013c), what leads a racial/ethnic group to experience anhedonia may be very different. Previous studies suggest the possibility of cultural variation in anhedonia based upon both the interpretation of “loss of interest” as an assessment (Sibitz et al., 2010b), and relatedly, cultural differences in baseline levels of pos- itive affect (Eid & Diener, 2001; Kanazawa, White, & Hampson, 2007; Mesquita & Karasawa, 2002). However, anhedonia may be difficult to predict based upon reported stressful life events and other assessments may be required to better understand the development of anhedonia. To determine the contribution of these factors, more research on both the development and context of anhedonia is needed.

Study Strengths and Limitations

These findings are bolstered by several study strengths. Our large sample size enabled us to stratify by race and ethnicity. Moreover, we were able to include APIs, a group that is not often examined in these types of analyses. Importantly, our findings highlight the domains that are salient to specific groups, an area that is difficult to assess and com- pare at a population level. We believe that our ability to conduct these analyses can greatly inform the design of assessments and preventions for PPD.

Limitations to this study include the reliance of broad racial/ethnic categories given that heterogeneous sub-ethnic groups may hold specif- ic cultural norms that present as a risk for postpartum depression. De- spite a large sample size, we had insufficient statistical power to apply a domain specific cumulative approach while stratifying by race/ethnic- ity, an approach in our previous work that accounts for both domains and cumulative stressors by developing eight different profiles of stress as PPD predictors (Liu & Tronick, 2013b). Such an approach would hone in on particular profiles (e.g., high financial + high relational + low physical) that are more or less predictive for the different racial/ethnic groups. We were also limited to a self-report of stressful life events dur- ing pregnancy, with no information on the frequency of such events or information on when the events occurred during pregnancy. Better re- call of stressful life events and identifying the chronicity and timing of stressful life events may be useful in predicting the likelihood of post- partum depressed mood and anhedonia. Relatedly, other moderating factors such as social support was not obtained in this sample; therefore, we are unable to determine any differences in social support within our sample and whether social support moderates or mediates the associa- tion between stressful life events and PPD symptoms. Finally, our out- come variables rely on one “yes/no” response for questions on each of the two core depression symptoms, using questions from common PPD screeners, such as the Patient Health Questionnaire. While this ap- proach may be externally valid, we are limited to the experience of only these two symptoms, without information about whether these experi- ences are distressing for the respondents.

Implications

Given the association between stressful life events during pregnancy and PPD symptoms, and the implications of this early life adversity on a broad range of children’s health across development (Kiernan & Huerta, 2008; Shonkoff, 2010). PPD detection should occur earlier, and at the very least, once a woman is pregnant. PPD screening can be improved beyond routine measures, including the use of core questions for de- pression (“During the past month, have you been bothered by feeling down, depressed, or hopeless; during the past month, have you been bothered by little interest or pleasure in doing things?”). Such questions may not be sufficiently sensitive for racial/ethnic groups. Clinicians that

work with pregnant women often inquire about well-being through open questions (e.g., “How are things going?”). Our data suggests that, at the minimum, simply focusing these inquiries based on stressful life domains may be informative in understanding PPD risk for minority groups, such as AAs and Hispanics. (“How are your relationships?” “How are things financially?”). As well, racial/ethnic minorities may be more willing to answer question about their life overall. These non- stigmatizing questions can supplement direct inquiry regarding de- pressed mood and loss of interest. Altogether, simple inquiries about stressful life events can potentially address health disparities for both racial/minority mothers and children.

Acknowledgment

The authors would like to acknowledge the Massachusetts Depart- ment of Public Health PRAMS Team, the CDC PRAMS Team, Program Services and Development Branch, Division of Reproductive Health, and Massachusetts State Representative Ellen Story and her staff. Finan- cial support during the preparation of this manuscript was provided through a Health Disparities Postgraduate Fellowship from the Harvard Medical School Office of Diversity and Community Partnership and the Commonwealth Research Center (SCDMH82101008006).

References

Arroll, B., Goodyear-Smith, F., Crengle, S., Gunn, J., Kerse, N., Fishman, T., ... Hatcher, S. (2010). Validation of PHQ-2 and PHQ-9 to screen for major depression in the primary care population. Annals of Family Medicine, 8(4), 348–353. http://dx.doi.org/10.1370/ afm.1139.

Bauer, A. M., Chen, C. -N., & Alegría, M. (2012). Associations of physical symptoms with perceived need for and use of mental health services among Latino and Asian Americans. Social Science & Medicine (1982), 75(6), 1128–1133. http://dx.doi.org/ 10.1016/j.socscimed.2012.05.004.

Beck, C. T. (2003). Recognizing and screening for postpartum depression in mothers of NICU infants. Advances in Neonatal Care: Official Journal of the National Association of Neonatal Nurses, 3(1), 37–46 (Retrieved from http://www.ncbi.nlm.nih.gov/ pubmed/12882180).

Beeghly, M., Olson, K. L., Weinberg, M. K., Pierre, S. C., Downey, N., & Tronick, E. Z. (2003). Prevalence, Stability, and Socio-Demographic Correlates of Depressive Symptoms in Black Mothers During the First 18 Months Postpartum. Maternal and Child Health Journal, 7(3), 157–168.

Broman, C. L. (1996). The health consequences of racial discrimination. Ethnicity and Disease, 6, 146–153.

Burchinal, M. R., Roberts, J. E., Hooper, S., & Zeisel, S. A. (2000). Cumulative risk and early cognitive development: A comparison of statistical risk models. Developmental Psychology, 36(6), 793–807.

Da Costa, D., Larouche, J., Dritsa, M., & Brender, W. (2000). Psychosocial correlates of prepartum and postpartum depressed mood. Journal of Affective Disorders, 59, 31–40.

Eid, M., & Diener, E. (2001). Norms for experiencing emotions in different cultures: inter- and intranational differences. Journal of Personality and Social Psychology, 81(5), 869–885.

Gavin, N. I., Gaynes, B. N., Lohr, K. N., Meltzer-Brody, S., Gartlehner, G., & Swinson, T. (2005). Perinatal depression: a systematic review of prevalence and incidence. Obstetrics and Gynecology, 106(5 Pt 1), 1071–1083 (Retrieved from http://www. ncbi.nlm.nih.gov/pubmed/16260528).

Gavin, A. R., Melville, J. L., Rue, T., Guo, Y., Dina, K. T., & Katon, W. J. (2011). Racial differ- ences in the prevalence of antenatal depression. General Hospital Psychiatry, 33(2), 87–93. http://dx.doi.org/10.1016/j.genhosppsych.2010.11.012.

Hatch, S. L., & Dohrenwend, B. P. (2007). Distribution of traumatic and other stressful life events by race/ethnicity, gender, SES and age: A review of the research. American Journal of Community Psychology, 40(3-4), 313–332. http://dx.doi.org/10.1007/ s10464-007-9134-z.

Hobfoll, S. E. (1998). Stress, Culture, and Community: The Psychology and Philosophy of Stress. New York City: Plenum Press.

Hobfoll, S. E., Ritter, C., Lavin, J., Hulsizer, M. R., & Cameron, R. P. (1995). Depression prev- alence and incidence among inner-city pregnant and postpartum women. Journal of Consulting and Clinical Psychology, 63(3), 445–453 (Retrieved from http://www. ncbi.nlm.nih.gov/pubmed/7608357).

Howell, E. A., Mora, P. A., Horowitz, C. R., & Leventhal, H. (2005). Racial and Ethnic Differ- ences in Factors Associated With Early Postpartum Depressive Symptoms. Race, 105(6), 1442–1450. http://dx.doi.org/10.1097/01.AOG.0000164050.34126.37.

Johnstone, S. J., Boyce, P. M., Hickey, A. R., Morris-Yatees, A. D., & Harris, M. G. (2001). Ob- stetric risk factors for postnatal depression in urban and rural community samples. The Australian and New Zealand Journal of Psychiatry, 35(1), 69–74 (Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11270460).

Kanazawa, A., White, P. M., & Hampson, S. E. (2007). Ethnic variation in depressive symp- toms in a community sample in Hawaii. Cultural Diversity and Ethnic Minority Psychology, 13, 35–44.

12 C.H. Liu et al. / Archives of Psychiatric Nursing 30 (2016) 7–12

Kendler, K. S., Myers, J., & Prescott, C. A. (2005). Sex differences in the relationship be- tween social support and risk for major depression: A longitudinal study of opposite-sex twin pairs. The American Journal of Psychiatry, 162(2), 250–256. http:// dx.doi.org/10.1176/appi.ajp.162.2.250.

Kiernan, K. E., & Huerta, M. C. (2008). Economic deprivation, maternal depression, parent- ing and children’s cognitive and emotional development in early childhood. The British Journal of Sociology, 59(4), 783–806. http://dx.doi.org/10.1111/j.1468-4446. 2008.00219.x.

Kroenke, K., Spitzer, R. L., & Williams, J. B. W. (2003). The Patient Health Questionnaire-2: validity of a two-item depression screener. Medical Care, 41(11), 1284–1292. http:// dx.doi.org/10.1097/01.MLR.0000093487.78664.3C.

Lancaster, C. A., Gold, K. J., Flynn, H. A., Yoo, H., Marcus, S. M., & Davis, M. M. (2010). Risk factors for depressive symptoms during pregnancy: A systematic review. American Journal of Obstetrics and Gynecology, 202(1), 5–14. http://dx.doi.org/10.1016/j.ajog. 2009.09.007.

Liu, C. H., & Tronick, E. (2013a). Prevalence and predictors of maternal postpartum de- pressed mood and anhedonia by race and ethnicity. Epidemiology and Psychiatric Sciences(August), 1–9. http://dx.doi.org/10.1017/S2045796013000413.

Liu, C. H., & Tronick, E. (2013b). Rates and predictors of postpartum depression by race and ethnicity: Results from the 2004–2007 New York City PRAMS survey (Pregnancy Risk Assessment Monitoring System). Maternal and Child Health Journal, 17(9), 1599–1610.

Liu, C. H., & Tronick, E. (2013c). Re-conceptualising prenatal life stressors in predicting post-partum depression: cumulative-, specific-, and domain-specific approaches to calculating risk. Paediatric and Perinatal Epidemiology, 27(5), 481–490. http://dx.doi. org/10.1111/ppe.12072.

Lu, M. C., & Chen, B. (2004). Racial and ethnic disparities in preterm birth: The role of stressful life events. American Journal of Obstetrics and Gynecology, 191(3), 691–699. http://dx.doi.org/10.1016/j.ajog.2004.04.018.

Mesquita, B., & Karasawa, M. (2002). Different emotional lives. Cognition & Emotion, 16(1), 127–141. http://dx.doi.org/10.1080/0269993014000176.

Myers, H. F. (2009). Ethnicity- and socio-economic status-related stresses in context: an integrative review and conceptual model. Journal of Behavioral Medicine, 32(1), 9–19. http://dx.doi.org/10.1007/s10865-008-9181-4.

Nilchaikovit, T., Hill, J. M., & Holland, J. C. (1993). The effects of culture on illness behavior and medical care: Asian and American differences. General Hospital Psychiatry, 15(1), 41–50 (Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8436284).

O’Hara, M. W., & Swain, A. M. (1996). Rates and risk of postpartum depression—a meta- analysis. International Review of Psychiatry, 8(1), 37–54. http://dx.doi.org/10.3109/ 09540269609037816.

Rich‐Edwards, J. W., Kleinman, K., Abrams, A., Harlow, B. L., McLaughlin, T. J., Joffe, H., & Gillman, M. W. (2006). Sociodemographic predictors of antenatal and postpartum

depressive symptoms among women in a medical group practice. Journal of Epidemiology & Community Health, 60, 221–227.

Ritter, C., Hobfoll, S. E., Lavin, J., Cameron, R. P., & Hulsizer, M. R. (2000). Stress, psychoso- cial resources, and depressive symptomatology during pregnancy in low-income, inner-city women. Health Psychology, 19(6), 576–585 (Retrieved from http://www. ncbi.nlm.nih.gov/pubmed/11129361).

Robertson, E., Grace, S., Wallington, T., & Stewart, D. E. (2004). Antenatal risk factors for postpartum depression: a synthesis of recent literature. General Hospital Psychiatry, 26, 289–295.

Shonkoff, J. P. (2010). Building a new biodevelopmental framework to guide the future of early childhood policy. Child Development, 81(1), 357–367. http://dx.doi.org/10.1111/ j.1467-8624.2009.01399.x.

Sibitz, I., Berger, P., Freidl, M., Topitz, A., Krautgartner, M., Spiegel, W., & Katschnig, H. (2010a). ICD-10 or DSM-IV? Anhedonia, fatigue and depressed mood as screening symptoms for diagnosing a current depressive episode in physically ill patients in general hospital. Journal of Affective Disorders, 126, 245–251.

Sibitz, I., Berger, P., Freidl, M., Topitz, A., Krautgartner, M., Spiegel, W., & Katschnig, H. (2010b). ICD-10 or DSM-IV? Anhedonia, fatigue and depressed mood as screening symptoms for diagnosing a current depressive episode in physically ill patients in general hospital. Journal of Affective Disorders, 126(1-2), 245–251. http://dx.doi.org/ 10.1016/j.jad.2010.03.023.

Thoits, P. A. (2010). Stress and health: major findings and policy implications. Journal of Health and Social Behavior(51 Suppl.), S41–S53. http://dx.doi.org/10.1177/ 0022146510383499.

Vega, W. A., Kolody, B., Valle, R., & Weir, J. (1991). Social networks, social support, and their relationship to depression among immigrant Mexican women. Human Organization, 50(2), 154–162.

Wei, G., Greaver, L. B., Marson, S. M., Herndon, C. H., & Rogers, J. (2008). Postpartum de- pression: racial differences and ethnic disparities in a tri-racial and bi-ethnic popula- tion. Maternal and Child Health Journal, 12(6), 699–707 (Retrieved from http://www. ncbi.nlm.nih.gov/pubmed/17955356).

Williams, D. R., Yu, Y., Jackson, J. S., & Anderson, N. B. (1997). Racial differences in physical and mental health: Socio-economic status, stress, and discrimination. Journal of Health Psychology, 2(3), 335–351.

Yonkers, K. A., Ramin, S. M., Rush, A. J., Navarrete, C. A., Carmody, T., March, D., ... Leveno, K. J. (2001). Onset and persistence of postpartum depression in an inner-city mater- nal health clinic system. The American Journal of Psychiatry, 158(11), 1856–1863 (Re- trieved from http://ajp.psychiatryonline.org/cgi/content/abstract/158/11/1856).

Zayas, L. H., Cunningham, M., McKee, M. D., & Jankowski, K. R. B. (2002). Depression and negative life events among pregnant African-American and Hispanic women. Women's Health Issues, 12(1), 16–22 (Retrieved from http://www.ncbi.nlm.nih.gov/ pubmed/11786288).