PTSD and postpartum mental health in a sample of Caucasian, Asian, and Pacific Islander women
- 433 Downloads
To better understand the role of posttraumatic stress disorder (PTSD) in postpartum health, this study investigates the relationship of PTSD and associated perinatal behavioral risk factors in a sample of Caucasian, Asian, and Pacific Islander women. As part of a larger longitudinal study, 54 women (18–35 years of age) were interviewed at their postpartum clinic visit for PTSD, anxiety, depression, and alcohol and substance use. PTSD and subclinical PTSD during the postpartum period were associated with behavioral health risks, with PTSD at the onset of pregnancy being a predictor of postpartum PTSD by a factor of three. Women with PTSD and subclinical PTSD were more likely to also experience stress (73%), anxiety (64%), and depression (73%) during the postpartum period compared to those without PTSD. No significant differences were found by ethnicity for postpartum PTSD, depression, or anxiety. Regardless of ethnicity or PTSD status, one in four women in the sample had a probable mental health disorder or risk behavior of some type during the postpartum period. Given the rates of associated mental health risks with PTSD, these findings suggest further research examining the fluctuations of PTSD symptomatology throughout each pregnancy trimester to determine its role as a potential mediator during the perinatal period. Further research is also needed to elucidate the role of ethnic or cultural differences in trauma and PTSD and perinatal health.
KeywordsPostpartum Subclinical PTSD Mental health Ethnicity
Relatively little attention has been given to the perinatal period in terms of risk for different kinds of mental disorders and describing risk factors from physical, psychological, and social stresses. Although disorders such as postpartum depression have gained greater attention as perinatal mental health issues (Brockington 2004; Dossett 2008; Wisner et al. 2006), less is known about the behavioral health risk factors associated with posttraumatic stress disorder (PTSD) and traumatic stress which are also likely to have an impact on postpartum health. PTSD has been linked to variety of behavioral health risk factors such as substance abuse and poor health care (Breslau et al. 1991; Schnurr and Jankowski 1999), which in turn, can impact perinatal health and potentially link to adverse birth outcomes such as preterm delivery, babies that are small or large for gestational age, or other complications (Born et al. 2006; Randall 2001; Rogal et al. 2007; Rosen et al. 2007; Russell and Skinner 1988; Seng et al. 2008). However, the relationship between PTSD and behavioral health during pregnancy and the postpartum period has not been fully explored.
PTSD has gained increasing attention as a widespread problem, however the potential implications for women’s mental and reproductive health have not been significantly addressed. Women are twice as likely to develop PTSD following a traumatic event compared to men (Kessler et al. 1995), with traumatic experiences often occurring prior to or during childbearing years (Breslau et al. 1991). PTSD prevalence rates are estimated to range from 9.7% to 12.3% for women having PTSD in their lifetime, and 4.6% to 5.2% of women with current PTSD (Kessler et al. 2005; Resnick et al. 1993). Further, women are exposed to interpersonal trauma (e.g., abuse or assault) at under-reported rates (Resnick et al. 1993). Although traumatic events and PTSD are still under-recognized risk factors in pregnancy (Mezey et al. 2005; Seng et al. 2001) and birth outcomes (Littleton et al. 2007; Mohler et al. 2008; Rosen et al. 2007; Seng. 2002), recent research suggests that PTSD may act as a mediator between traumatic life events and behavioral health risks during pregnancy, including alcohol, smoking, substance abuse, depression, anxiety, and stress (Morland et al. 2007; Seng. 2002).
Of anxiety disorders that have been found to be frequent during pregnancy and after delivery (Brockington et al. 2006; Wenzel et al. 2003, 2005), rates of obsessive–compulsive disorder and generalized anxiety disorder have been reported as higher in postpartum women than in the general population (Ross and McLean 2006). However, relatively little is known about PTSD during pregnancy, and in particular, the postpartum period (Brockington et al. 2006; Zaers et al. 2008). The existing literature on postpartum PTSD typically focuses on PTSD related to traumatic childbirth (Cohen et al. 2004; Maggioni et al. 2006; Olde et al. 2006; Reynolds 1997), and does not address the potential significance of fluctuating (Born et al. 2006) or persisting PTSD symptomatology, which in women may be more commonly related to past or current interpersonal trauma exposure (Duncan et al. 1996; Mezey et al. 2005), throughout pregnancy and during the postpartum period. For women with re-emergence of PTSD symptomatology (Born et al. 2006) or exposure to further trauma during pregnancy (Bacchus et al. 2004), associated risk behaviors (alcohol, smoking, substance use) and other psychological comorbidities (depression and anxiety) may continue to persist or become exacerbated through the postpartum period, a time during which women may be vulnerable to mental health issues without adequate screening and treatment. If behavioral risk and mental health factors are present at the onset of pregnancy, it is likely that their role in maternal behavioral health into the postnatal period will, in turn, have important implications for child development (Brand et al. 2006; Moehler et al. 2006).
Another aspect that highlights the gap in the literature is that the prevalence of perinatal PTSD varies by cultural context. Studies with ethnically diverse populations have included African–American and Hispanic groups (Pole et al. 2005; Rodriguez et al. 2008; Smith et al. 2006), however, there is less data on Asian and Pacific Islander women on perinatal mental health (Goebert et al. 2007). Data from a longitudinal study (Morland et al. 2007) show over 60% prevalence of abuse trauma in a sample of Asian, Pacific Islander, and Caucasian women seeking prenatal care in an ethnically diverse region of Hawai‘i. Both Asian (19%) and Pacific Islander (24%) women had higher rates of PTSD and subclinical PTSD compared to Caucasian women (9%) at the beginning of their pregnancy, which were associated with behavioral health risks such as alcohol use and smoking during pregnancy, substance use, depression, anxiety, and poor prenatal care.
To better understand the role of posttraumatic stress disorder in postpartum health, this study investigates the relationship of PTSD and associated perinatal behavioral risk factors in a sample of Caucasian, Asian, and Pacific Islander women. This study explores whether PTSD, present at the onset of pregnancy, is predictive for PTSD during the postnatal period. It is also hypothesized that postpartum PTSD or subclinical PTSD is associated with other mental health disorders such as depression and anxiety, as well as with substance use. Finally, this study also seeks to examine potential ethnic differences for Asian, Pacific Islander, and Caucasian women in PTSD and associated mental and behavioral health risks.
Postpartum data are presented from a larger longitudinal study of perinatal behavioral health (Goebert et al. 2007) approved by the local Institutional Review Board, that was conducted over a period of 4 years (2002–2006). In the larger study, 122 women were assessed on four occasions during the course of their pregnancy: (1) at their initial prenatal visit, usually during first trimester, (2) at second trimester, (3) at third trimester, and (4) at postpartum. From the larger study sample, 54 women of Asian (37%), Caucasian (22%), and Native Hawaiian or other Pacific Island (41%) ethnicity were interviewed during the postpartum period. The study participants were from among those who received prenatal care at an outpatient obstetric-gynecology (ob/gyn) clinic or private physicians’ offices associated with a local community medical center specializing in the care of women and children. Women ages 18–35 presenting for their initial prenatal care visit were invited to participate in this study. The present sample includes data from women interviewed at their postpartum clinic visit (between 4 and 8 weeks).
Participants were interviewed with informed consent during their prenatal and postpartum visits in a private area of the waiting room of the clinic or physician’s office. The interview was conducted in English, and lasted approximately 30 min. Interview questions at the initial screening included: demographic information (age, ethnicity, marital status, education, and household income), substance use (alcohol, smoking, and other illicit drug use), general stress, history of trauma, PTSD symptomatology, anxiety, depression, and social support. The interview was repeated with a briefer questionnaire (approximately 20 min) during their postpartum visit. Participants were compensated with a $5 gift certificate to a local retailer for the initial interview, and a $10 gift certificate for the postpartum interview.
Trauma and PTSD
The Traumatic Life Events Questionnaire (TLEQ; Kubany et al. 2000) was used to assess the various types of trauma that women had experienced during their lifetime, including interim trauma during pregnancy and postpartum. Categories of trauma included interpersonal violence such as domestic abuse, sexual assault, and child abuse. The TLEQ also assessed the participants’ helplessness or horror in response to the endorsed trauma exposure, and the event source of significant distress.
Lifetime and current PTSD symptoms following the reported trauma exposure were measured using the PTSD Checklist-Civilian version (PCL-C; Weathers and Litz 1994). The PCL-C includes items (reexperiencing, avoidance, arousal) based on the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV; 1994) criteria used to determine PTSD diagnosis. Item responses, reflecting a range of symptom intensity, were used to also calculate a total clinical symptomatology score. In order to capture a potentially significant subgroup of individuals with subclinical PTSD (Breslau et al. 2004; Manne et al. 1998; Morland et al. 2007; Stein et al. 1997; Yarvis and Schiess 2008) that may also have similar levels of impairment or relationship to associated mental health and behavioral risk factors, a subclinical PTSD category was included. There is some variation on the definition of subclinical or partial PTSD and the corresponding criteria used (Breslau et al. 2004; Morland et al. 2007; Stein et al. 1997; Wang et al. 2000), however, to be consistent with the larger study, we elected to define subclinical PTSD as meeting the required number of symptoms for each of the categories of PTSD, allowing for a lesser intensity (Morland et al. 2007).
Smoking, alcohol, and drug use
Cigarette, alcohol, and illicit drug use was obtained from participant self-report. Participants were asked how many cigarettes per day were smoked; how frequently they drank alcohol (beer, wine or liquor), amount (average when drinking), and binge drinking (number of times they had had five or more drinks on one occasion); frequency of illicit drug use (none, 1–2×/month, 1–2×/week, 3–5×/week, or daily).
The participants were also screened for problematic alcohol use using the TWEAK (Tolerance, Worry, Eye-opener, Amnesia, & Cut down drinking) (Russell and Bigler 1979). The TWEAK, a 5-item self-report measure, has been shown to be a good screening instrument for the detection of harmful and dependent drinking, and has been validated among women (Cherpitel 1995; D’Onofrio et al. 1998), and in particular, obstetric and gynecologic outpatients (Russell et al. 1994, 1996; Russell and Skinner 1988). A score of 2 was used to indicate “at risk” drinkers, while scores of 3 or higher were labeled as “probable” (Chan et al. 1993).
The Center for Epidemiologic Studies Depression Scale (CES-D; Radloff 1977), a 20-item self-report screening instrument, was used to measure symptoms of depression in the past 7 days. The CES-D has been used as a screener for depression (score of 16 or higher) in women during and after pregnancy (Marcus et al. 2003; Mosack and Shore 2006).
Levels of state anxiety at each assessment were measured using the 20-item State-Trait Anxiety Inventory (STAI; Spielberger et al. 1983). The STAI is the most widely used self-report measure of anxiety in adults. Cutoff scores of 40 or higher were considered to be screen positive and at risk for an anxiety disorder (Rondo et al. 2004).
General stress was measured using a short 10-item version of the Perceived Stress Scale (PSS; Cohen et al. 1983). The self-report items ask respondents to appraise general and nonspecific distress.
Social support was measured with an abbreviated 6-item version of the 27 self-report item Social Support Questionnaire (SSQ; Sarason et al. 1983). Each item assesses: (1) the number of available others the individual feels s/he can turn to in times of need in a variety of situations, and (2) the individual’s degree of satisfaction with the perceived support available in that particular situation.
Descriptive and bivariate statistical analyses, including chi-square tests, were conducted to compare demographic and behavioral health characteristics of the postpartum sample to the larger study. Chi-square analyses were also used to compare postpartum PTSD and other behavioral health problems (depression, anxiety, problems with alcohol) by ethnicity. Logistic regression was used to determine predictors for postpartum PTSD. Chi-square analyses were used to examine associations between postpartum PTSD with depression, anxiety, and substance use. All analyses were performed using SPSS 16.0 statistical software.
Because of the relatively smaller number of women (n = 54) in the postpartum sample, preliminary analyses were conducted to test representativeness of demographic and behavioral risk characteristics to the larger study sample, as assessed at the initial prenatal visit.
Comparison of demographic characteristics and parity for women with and without postpartum interviews
No postpartum interview
Age (N = 110)a
M = 27.54
M = 26.85
t= 0.739; p = 0.459, two-tailed
SD = 4.230
SD = 5.381
Ethnicity (N = 107)
χ2  = 0.082; p = 0.960
Marital Status (N = 107)
χ2  = 0.255; p = 0.613
Education (N = 107)
χ2  = 2.216; p = 0.330
High School or less
College degree or higher
Household income (N = 77)
χ2  = 0.429; p = 0.429
$40,000 or less
Parity (N = 97)
χ2  = 0.903; p = 0.342
Rates of PTSD and other behavioral health screening during the postpartum period for Caucasian, Asian, and Pacific Islander women
Total (n = 54)
Caucasian (n = 12)
Asian (n = 20)
Pacific Islander (n = 22)
χ2 ; p
7.316; p = 0.120
5.794; p = 0.215
4.911; p = 0.297
5.746; p = 0.219
12.293; p = 0.015*
For subsequent analyses, the PTSD and subclinical PTSD subgroups were collapsed into a single category. Approximately 62% of the postpartum PTSD cases that were comorbid with postpartum depression were also depressed prenatally (χ2  = 3.909; p = 0.048), and about 83% of postpartum PTSD cases comorbid with an anxiety disorder had previously presented at the initial assessment with an anxiety disorder (χ2  = 7.824; p = 0.005). With prenatal PTSD, prenatal depression, prenatal anxiety, and trauma type as predictor variables for postpartum PTSD, a logistic regression analysis was performed. A total of 43 cases were analyzed, and the model significantly predicted postpartum PTSD (omnibus χ2  = 17.834; p = 0.001). The model coefficients revealed that women with prenatal PTSD at the initial assessment were more than 3 times as likely to have postpartum PTSD (OR 3.648; 95% CI 1.477, 998.808; p = 0.028). A trend was also found for trauma type, for both recent and old traumatic events which were the root source, as a predictor for postpartum PTSD. Women who experienced and reported trauma that was related to interpersonal violence as distressing were twice as likely to have postpartum PTSD (OR 2.021; 95% CI 0.973, 58.539; p = 0.053). However, neither prenatal anxiety (OR 1.924; 95% CI 0.843, 55.639; p = 0.072) nor prenatal depression at early pregnancy (OR −2.543; 95% CI 0.003, 2.349; p = 0.142) were predictive for postpartum PTSD.
Rates of comorbid mental health risks and substance use by PTSD during the postpartum period
Total (n = 54)
No PTSD (n = 43)
PTSD or subclinical PTSD (n = 11)
χ2 ; p
17.89; p = 0.000**
15.942; p = 0.000**
23.345; p = 0.000**
0.069; p = 0.793
3.892; p = 0.049*
0.057; p = 0.811
This study is one of few in the existing literature that includes research on PTSD and pregnancy among women of diverse ethnic backgrounds, specifically Asians and Pacific Islanders (Goebert et al. 2007; Kubany et al. 1996, 2003, 2004; Morland et al. 2007). Although rates of PTSD, subclinical PTSD, depression, or anxiety during the postpartum period did not vary significantly by ethnicity, Pacific Islander women showed highest overall rates for probable behavioral health problems that included a disorder of any type. It is important to note that the absent rates of mental health problems for Caucasian women should not be taken to suggest that a particular ethnic group is unaffected by mental health issues as it is unlikely that prevalence is zero within the population. Regardless of ethnicity, one in four women had some kind of mental health disorder, and 19% were at risk, particularly for depression and PTSD. These rates support the growing literature that the postpartum period may be a time of vulnerability to mental health issues that may be, in fact, preexisting but undiagnosed and untreated (Loveland Cook et al. 2004).
The role of PTSD in antepartum and postpartum periods is not yet well understood. Consistent with our hypotheses, women with PTSD and subclinical PTSD during the postpartum period were more likely to also have depression, anxiety, higher perceived stress, and binge drink compared to women without any PTSD or subclinical PTSD. However, we did not find comparable rates of PTSD in our sample of postpartum women (roughly 2% PTSD, and 19% subclinical PTSD) relative to the larger study (16% PTSD, and 23% subclinical PTSD) (Morland et al. 2007).
Considerations regarding the subclinical PTSD subgroup include the possibility that the category may have been an artifact comprised of individuals that had a temporary reduction in symptom severity during the trimester interview. Another factor that may be important to consider is the influence of cultural responding of overall symptom endorsement and symptom intensity (Wang et al. 2000). In traditional Asian cultures, such as those which can still be seen reflected in current cultural upbringing in Hawai‘i, the endorsement of psychological distress or symptom severity may be underreported. Additionally, due to small sample size, those with PTSD and subclinical PTSD were collapsed into a single category for subsequent analyses.
From our model, the best predictor for postpartum PTSD was prenatal PTSD as assessed in early pregnancy, however, we also included and found a trend for interpersonal violence types of trauma related to the PTSD as an additional predictor. A closer look at the proportion of postpartum PTSD cases showed almost two-thirds of the women endorsed the same traumatic event as reported in the initial assessment as the most distressing, and approximately one-third of the cases were related to new trauma exposure around the time of delivery. Of the types of traumatic events, regardless of previous or new exposure, a majority (73%) were related to interpersonal violence (i.e., childhood sexual abuse, intimate partner violence, family violence, sexual assault) compared to 27% related to other types of trauma (i.e., death or illness of a loved one). None of the postpartum PTSD cases were related to birth trauma. This corroborates the need to further explore traumatic stressors that may be more related to interpersonal violence for women of child-bearing age.
Although the women interviewed at postpartum did not vary on demographic and behavioral health characteristics from women who were not assessed, the small sample size limits the generalizability of these findings. It is plausible that the women who returned for the postpartum assessments were different from those who did not on a number of other important mental health issues. Given that rates of comorbidities with PTSD, depression, and other anxiety disorders are reported as high (Loveland Cook et al. 2004), we did not utilize diagnostic screening measures, because of feasibility, for other psychiatric illnesses (bipolar disorder, OCD, panic disorder) that may be important to consider.
Due to the limitations of the study, clinical implications are modest, however, a potential recommendation for perinatal providers who serve a high-risk population and are already implementing behavioral health screening may include supplementation of existing measures with assessments for trauma and PTSD, and follow up of women who screen positive for PTSD in early pregnancy. Prenatal, postnatal and postpartum care allows for multiple opportunities for screening (Borri et al. 2008; Reck et al. 2008) which may facilitate improved outcomes for women during the perinatal period (Seng 2003).
Future directions of research from this study include examining fluctuations of PTSD symptomatology throughout each pregnancy trimester. Based on the currently available literature (Morland et al. 2007; Seng 2002), it is yet unclear whether PTSD acts as a mediator for behavioral health risks, throughout the prenatal and into the postpartum periods. All but two of the postpartum PTSD cases in our study were symptomatic prenatally, however, the incidence of new PTSD during postpartum would be difficult to verify without also examining other trimester data. Sequential analyses of trimester data may shed light on further understanding the variation of PTSD during and after pregnancy (Smith et al. 2006).
Potential implications of PTSD-related health risk behaviors may impact birth outcomes and long-term maternal mental health. More longitudinal approaches in research well into the postpartum period (6–12 months) would allow continued follow up of women who screen positive for behavioral health risk factors in order to determine the subsequent impact of maternal mental health on early child development.
Finally, although there were no significant differences by ethnicity in individual perinatal behavioral health risks in our postpartum sample, further research is needed to elucidate the role of ethnic or cultural differences in trauma and PTSD and perinatal health. In order to address the gap on mental health disparities of women from ethnic minority and disadvantaged populations during their childbearing years, more research is clearly needed to better understand who may be at higher risk for postpartum health issues.
This research was supported, in part, by a Research Centers in Minority Institutions Award, P20 RR11091, from the National Center for Research Resources (NCRR), National Institutes of Health. Dissemination is supported by a National Institute on Alcohol Abuse and Alcoholism (NIAAA) and National Center on Minority Health and Disparities (NCMHD) Award (5U01AA014289-03). The contents of this paper are solely the responsibility of the authors and do not necessarily represent the official views of NIAAA, NCMHD, and NCCR. The investigators thank the women and staff from the clinics that participated in the study.
- Borri C, Mauri M, Oppo A, Banti S, Rambelli C, Ramacciotti D et al (2008) Axis I psychopathology and functional impairment at the third month of pregnancy: results from the perinatal depression-research and screening unit (PND-ReScU) study. J Clin Psychiatry 69(10):1617–1624CrossRefPubMedGoogle Scholar
- Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) (1994) (4th ed). American Psychiatric Association, Washington, DCGoogle Scholar
- D’Onofrio G, Bernstein E, Bernstein J, Woolard RH, Brewer PA, Craig SA et al (1998) Patients with alcohol problems in the emergency department, part 1: improving detection. SAEM substance abuse task force. Society for academic emergency medicine. Acad Emerg Med 5(12):1200–1209PubMedCrossRefGoogle Scholar
- Spielberger CD, Gorsuch RL, Lushene R, Vagg PR, Jacobs GA (1983) Manual for the state-trait anxiety inventory. Consulting Psychologists, Palo AltoGoogle Scholar
- Weathers FW, Litz BT (1994) Psychometric properties of the Clinician-Administered PSTD Scale, CAPS-1. PTSD Research Quarterly 5:2–6Google Scholar