Experiences of Intimate Partner and Neighborhood Violence and Their Association With Mental Health in Pregnant Women

Abstract

Both intimate partner violence and neighborhood crime have been associated with worse mental health outcomes, but less is known about cumulative effects. This association was studied in a sample of pregnant women who were enrolled in a study of disaster exposure, prenatal care, and mental and physical health outcomes between 2010 and 2012. Women were interviewed about their exposure to intimate partner violence and perceptions of neighborhood safety, crime, and disorder. Main study outcomes included symptoms of poor mental health; including depression, pregnancy-specific anxiety (PA), and posttraumatic stress disorder (PTSD). Logistic regression was used to examine predictors of mental health with adjustment for confounders. Women who experienced high levels of intimate partner violence and perceived neighborhood violence had increased odds of probable depression in individual models. Weighted high cumulative (intimate partner and neighborhood) experiences of violence were also associated with increased odds of having probable depression when compared with those with low violence. Weighed high cumulative violence was also associated with increased odds of PTSD. This study provides additional evidence that cumulative exposure to violence is associated with poorer mental health in pregnant women.

Keywords

violence neighborhood intimate partner violence mental health pregnancy

Introduction

Women of childbearing age are at increased risk for intimate partner violence (IPV; El Kady, Gilbert, Xing, & Smith, 2005; Silverman, Decker, Reed, & Raj, 2006), which commonly occurs in pregnancy (James, Brody, & Hamilton, 2013), and often results in adverse mental health outcomes and serious physical injury (McFarlane, Campbell, Sharps, & Watson, 2002). IPV may include emotional, physical, or sexual violence, and mental health outcomes become more severe with more than one type of IPV (Lagdon, Armour, & Stringer, 2014). IPV is consistently linked to depression and other mental health issues in pregnancy (Urquia, O’Campo, Heaman, Janssen, & Thiessen, 2011), and increased risk for adverse birth outcomes (Shah, Shah, & Knowledge Synthesis Group on Determinants of Preterm/LBW Births, 2010). Women exposed to violence in pregnancy are also at increased risk of serious injury or homicide (McFarlane et al., 2002). This risk is even higher for African American women compared with White women (Chang, Berg, Saltzman, & Herndon, 2005).

Neighborhood violence is an additional risk factor for poor mental health (Aisenberg & Herrenkohl, 2008; Egan, Tannahill, Petticrew, & Thomas, 2008; Kelly & Hall, 2010; Wilson-Genderson & Pruchno, 2013). Witnessing or experiencing neighborhood violence has been associated with depression (Clark et al., 2008; Gillespie et al., 2009; Tracy, Morgenstern, Zivin, Aiello, & Galea, 2014), posttraumatic stress disorder (PTSD; Goldmann et al., 2011), and anxiety (Casciano & Massey, 2012) in studies of urban nonpregnant adults. Neighborhood violence can also be a cause of increased stress and fraying of social ties. High levels of perceived crime have been associated with more psychological distress (Giurgescu et al., 2015), lack of social support, racism, and depression (Messer, Maxson, & Miranda, 2013). Low-income women are disproportionately exposed to neighborhood crime and violence (Pinchevsky & Wright, 2012), and pregnancy is an especially vulnerable time in a woman’s life. Few studies on neighborhood violence, however, have focused on women (Aisenberg & Herrenkohl, 2008), and none were found that included pregnant women.

There is also a paucity of published research on the cumulative effects of violence on pregnant women at both the individual and neighborhood levels. Exposure to multiple kinds of violence is not uncommon, and minority women disproportionately reside in communities where stress, racism, and violence are daily struggles (Amaro, Vega, & Valencia, 2001; Taylor & Holden, 2009). Only three studies were identified that investigated cumulative effects of violence at both the individual and neighborhood levels, results were inconsistent, and none of these included pregnant women in their study populations (Bogat et al., 2005; Brown, Hill, & Lambert, 2005; Clark et al., 2008). In one study, after adjustment for IPV, witnessing neighborhood violence was associated with increased odds of high depressive symptoms compared with those who never witnessed crime in their communities (odds ratio [OR] = 2.6, 95% confidence interval [CI] = [1.4, 4.9]; Clark et al., 2008). Others reported increased exposure to either IPV or neighborhood violence was associated with increased PTSD symptoms, and that additive interaction was present (as violence increased, so did trauma symptoms; Brown et al., 2005). Finally, Bogat et al. (2005) examined the association of cumulative violence, anxiety, depression, and PTSD using objective measures of geographic information systems (GIS) neighborhood crime data. They reported that when community violence was present without IPV, there were no adverse mental health outcomes; however, IPV alone was related to mental health, regardless of objective measures of neighborhood violence (Bogat et al., 2005).

The community context of violence and the long-term effects of natural disasters that contribute to violence must also be considered. Poor, urban communities are disproportionately affected by disasters and among the slowest to recover (Davidson, Price, McCauley, & Ruggiero, 2013; Fothergill, Maestas, & Darlington, 1999; Phillips, Thomas, Fothergill, & Blinn-Pike, 2010). New Orleans is one example of a city that has been plagued with violence as well as disaster, as Hurricane Katrina hit in 2005. There is some evidence that intimate partner violence rises after disaster (Harville, Taylor, Tesfai, Xu, & Buekens, 2011), though IPV levels were comparable to national levels (James et al., 2013). Overall crime statistics indicate a dip in crime in that city immediately following the disaster, then an increase afterwards, with a leveling-off in the last few years (Federal Bureau of Investigation, 2012). Nonetheless, New Orleans remains a city with a high violent crime rate, and crime is a constant concern for residents (CommonHealth ACTION, 2009). Therefore, social disempowerment, poverty, unemployment, and other factors related to IPV and neighborhood violence may be more important considerations for maternal and child health than simply living in a postdisaster recovery area.

Therefore, the cumulative effect of lack of safety within and outside the home for pregnant women has not been studied extensively. The purpose of this analysis was to examine how perceptions of crime and violence, both intimate partner and neighborhood-wide, as well as cumulative effects of both types of violence predicted adverse mental health outcomes in a cohort of New Orleans pregnant women.

Method

Women (N = 398) from the metropolitan New Orleans area were recruited for the GUMBO (Growing up Moms and Babies in new Orleans) study from 2010 to 2012. The objective of this cross-sectional study was to investigate how models of prenatal care, self-care strategies, and stress and hurricane recovery influence mental and physical health during pregnancy. Data were collected between 5 and 7 years after Hurricane Katrina. Participants were recruited from seven sites, including private prenatal clinics, University hospital-affiliated clinics, and prenatal and childbirth classes. Women were approached in clinic waiting rooms and from prenatal classes by trained data collectors at patients’ scheduled appointments. Eligible women spoke either English or Spanish, were between 18 and 45 years of age, had an established prenatal care provider (at least three visits), and were currently living in the Greater New Orleans area. Women were asked about exposure to Hurricane Katrina, which occurred in 2005, but moving to New Orleans after the storm did not exclude them from study participation, as the main exposure of interest in the original study was the effects of living in the postdisaster recovery environment. Informed consent was obtained in the preferred language of the client, and Spanish interviews were conducted by fully bilingual and bicultural data collectors. Participants’ responses to survey questions were recorded on paper questionnaires. The study was carried out according to protocol and institutional review board (IRB) approval was obtained from affiliated Universities and hospitals. Three women were missing data on crime or race, leaving 398 for analysis.

Exposure Assessment

The main exposures of interest were intimate partner violence and neighborhood violence. Intimate partner violence was assessed using items from the Centers for Disease Control and Prevention’s Pregnancy Risk Assessment Monitoring System (PRAMS) questionnaire (2009). Physical and emotional abuse during pregnancy was assessed using six “yes/no” questions. Three questions addressed physical abuse (whether their partner hit, hurt her, or forced sexual activity), and three questions addressed emotional abuse and issues of control (whether their partner threatened, frightened, or tried to control her). Positive responses of abuse were summed for each domain (physical and emotional abuse), and ranged from zero to three, with 0 = none, 1 indicator = some, 2 or more yes responses = high.

Perception of neighborhood violence, crime, and safety were also assessed. Women were asked to rate their perceptions of life in the city and expectations for the city’s future. The majority of the questions were taken from the Kaiser Family Foundation (2007) survey “Giving Voice to the People of New Orleans.” Questions on crime were taken from both general community surveys (e.g., How satisfied are you with police services? SOSNA/NAC/CDC Public Safety Committee, n.d.) and previous studies of neighborhood influences on well-being (Ajrouch, Reisine, Lim, Sohn, & Ismail, 2010; Earls, Brooks-Gunn, Raudenbush, & Sampson, 1994). Additional questions asked, “If you witnessed a serious crime, would you report it to the police?” and “Have you bought a weapon for your own or your family’s protection?” (Questions and response options are provided in Figure 1).

Figure 1.

Crime and safety questionnaire items.

A cumulative violence exposure index was then created to measure overall lack of safety. To achieve this, we first performed factor analysis on the six questions for intimate partner violence and the six questions for neighborhood violence. Intimate partner violence indicated two factors, consistent with the scale construction: emotional and physical violence. Overall intimate partner violence was summed across all six questions and reported on the same scale as the individual violence measures (0 = none, 1 indicator = some, 2 or more yes responses = high). Three factors were indicated for neighborhood violence: “Perceived neighborhood safety” (Figure 1: CS1-3 and 5), “Reporting” (CS4), and “Weapon ownership” (CS6). Weapon possession was unrelated to any of the outcomes studied as well as minimally related to other indicators of violence exposure, and was excluded from further analysis. “Perceived neighborhood safety” was created by dichotomizing each of the contributing variables to high/low, then summing the four questions to develop a total scale. Total neighborhood violence was created by weighting the “Perceived neighborhood safety” and “Reporting” factors equally, then summing and creating tertiles (equal weighting). The cumulative violence indicator was then calculated by summing the average score on the neighborhood and the intimate partner violence scales. A weighted cumulative violence indicator was also created by weighting the beta coefficient associated with the total neighborhood and total intimate partner violence scales. Finally, scores for each scale were converted to tertiles (low, medium, high).

Outcome Assessment

The primary study outcomes were depression, PA, and PTSD. The Edinburgh Postnatal Depression Scale (EDS) was used to assess symptoms of depression and has been validated for use in pregnancy in English and Spanish (Alvarado-Esquivel, Sifuentes-Alvarez, Salas-Martinez, & Martinez-Garcia, 2006), with a Cronbach’s alpha of .76 (Adouard, Glangeaud-Freudenthal, & Golse, 2005) and .85 (Adewuya, Ola, Dada, & Fasoto, 2006) in two diverse populations. The EDS is not a diagnostic measure but instead a screening tool that indicates “probable” or “likely” depression. Probable depression was defined as EDS score greater than 12 (Matthey, Henshaw, Elliott, & Barnett, 2006). PA was assessed using the Revised Prenatal Distress Questionnaire (Yali & Lobel, 1999) and was dichotomized (high score >17; top quintile) as others have done (Clements & Bailey, 2010) to identify the women with the highest prenatal anxiety. Cronbach’s alphas were reported between .80 and .81 among diverse groups internationally (Alderdice, Lynn, & Lobel, 2012). The PTSD checklist (PCL-S) is the posttraumatic checklist, which asks about symptoms related to a stressful experience. Ventureyra, Yao, Cottraux, Note, and De Mey-Guillard (2002) reported a Cronbach’s alpha of .86 for the PCL-S (Ventureyra et al., 2002). PTSD symptoms were dichotomized at 50 (Andrykowski, Cordova, Studts, & Miller, 1998).

We utilized the validated Spanish language versions of the EDS and PCL-S; the PCL-S in Spanish was found to be functionally equivalent to the English (Miles, Marshall, & Schell, 2008). The Revised Prenatal Distress Questionnaire was translated into Spanish, as no validated instrument was available. All instruments in Spanish were then examined for discrepancies with the original English versions by two independent, fully bilingual reviewers. We also calculated the reliability for the EDS, PA, and PCL-S instruments in our sample, and alpha ranged from .86 to .92. Among participants who only spoke Spanish, the alpha coefficient was .83, .71, and .92, respectively (n = 57).

Statistical Analysis

The relationship between intimate partner violence and neighborhood violence with mental health outcomes was assessed as follows. First, descriptive data and frequencies were computed, and then bivariate associations were tested using chi-square and Fisher’s exact tests. Second, unadjusted and adjusted multivariable logistic regression was employed to examine relationships between each aspect of intimate partner violence and neighborhood violence as well as the overall lack of safety variable was examined as a predictor of mental health outcomes. Third, both intimate partner violence and neighborhood violence were included in the same models. Finally, we calculated a cumulative measure of total violence, which incorporated both intimate partner and neighborhood violence to examine effects on mental health. Potential confounders (age, race, education, marital status, and smoking) were identified by a priori knowledge of the literature and included in multivariable models. Smoking during pregnancy was assessed via one question, “What is the average number of cigarettes that you smoke each day?” Responses were then categorized into yes/no groups for smoking during pregnancy. Missing data for covariates was dealt with using multiple imputation; most commonly missing was income for 6% (18) of participants. Due to a questionnaire error, for some participants, one question on the EDS was omitted and another was repeated. The mean was imputed based on the other EDS questions for these participants (n = 89). All analyses were conducted using SAS 9.3 (Cary, NC).

Results

The study population was largely in their 20s, majority African American, unmarried, and low-income (Table 1). The median and mean gestational age at interview was 31 weeks. Twelve percent of women interviewed reported intimate partner violence. The majority of respondents said that their neighborhoods were about the same or safer than last year. Nearly one third of participants stated that they were somewhat or very dissatisfied with police protection in their neighborhoods and that muggings occurred fairly or very often. Almost a third of the women in this sample had probable depression (EDS > 12), while fewer had high PA (> 17 and less commonly, symptoms of PTSD (PCL .50). All study participants lived in the metropolitan area within 5 years of the Hurricane Katrina’s landfall. Approximately 10% of the women who did not report living in New Orleans before Hurricane Katrina reported some degree of exposure to the storm, most likely due to living in other affected areas, or because friends or relatives were affected. In addition, there was no interaction found for main outcomes of interest by hurricane-survival status (data not shown).

Table 1.

Study Population.

	N	%
Age category
1 = <20	62	15.6
2 = >20-25	119	29.9
3 = >25-30	116	29.2
4 = >30	101	25.4
Race/ethnicity
White non-Hispanic	105	26.4
Black non-Hispanic	227	57.0
Hispanic	66	16.6
Education
Less than high school	89	22.5
High school	111	28.0
Greater than high school	196	49.5
Relationship status
Married	111	28.2
Living with partner	115	29.2
Single, divorced, or widowed	168	42.6
Annual income
<US$15,000	193	51.7
US$15,000-US$30,000	76	20.4
>US$30,000	104	27.9
Parity
Primipara	236	60.2
Multipara	156	39.8
Smoke
Yes	37	9.3
No	360	90.7
How quickly returned to New Orleans
<1 year	162	40.7
1-2 years	120	30.2
3-4 years	68	17.1
5+ years	48	12.1
Lived in New Orleans prior to Katrina
Yes	277	69.6
No	121	30.4
Serious experiences of the hurricane
Three or more	72	18.1
Less than three	326	81.9

Note. Percentages may not sum to 100 due to missing data.

Unadjusted and multivariable (adjusted) logistic regression was conducted next to investigate the associations between intimate partner violence and mental health outcomes (Table 2). Women who experienced some and high physical violence had higher odds of probable depression compared with women who reported no physical violence in both unadjusted and adjusted models. Similarly, women who reported some or high emotional intimate partner violence were more likely to have probable depression than women who did not report emotional violence. Women who had high levels of total intimate partner violence were more likely to have probable depression, in both unadjusted and adjusted models. Those who experienced some or high levels of physical intimate partner violence had significantly higher odds of PTSD compared with those with none. Women who reported high levels of emotional intimate partner violence also had significantly higher odds of PTSD. Total intimate partner violence was significantly associated with PTSD for women experiencing high violence compared with none. Total intimate partner violence was not significantly associated with PA; however, women who reported some physical intimate partner violence were more likely to have PA than those with none.

Table 2.

Unadjusted and Adjusted^a Multivariable Logistic Regression of Intimate Partner Violence and Mental Health Outcomes (N = 398).

			Probable Depression		PTSD		Pregnancy-Specific Anxiety
			Unadjusted	Adjusted	Unadjusted	Adjusted	Unadjusted	Adjusted
	N	%	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)
Overall intimate partner violence^b
None (0)	344	86.7	Reference	Reference	Reference	Reference	Reference	Reference
Some (1)	32	8.1	4.45 [2.11, 9.40]	3.92 [1.71, 9.02]	2.57 [0.90, 7.30]	1.75 [0.56, 5.46]	2.39 [1.07, 5.32]	2.15 [0.89, 5.20]
High (≥2)	21	5.3	9.75 [3.47, 27.41]	13.97 [4.43, 44.07]	5.55 [1.97, 15.65]	6.61 [1.90, 22.95]	1.64 [0.58, 4.67]	1.69 [0.54, 5.29]
Physical intimate partner violence^c
None (0)	365	91.9	Reference	Reference	Reference	Reference	Reference	Reference
Some (1)	21	5.3	4.49 [1.81, 11.16]	4.78 [1.76, 13.00]	5.41 [1.93, 15.15]	4.83 [1.48, 15.76]	3.19 [1.27, 8.04]	3.06 [1.13, 8.27]
High (≥2)	11	2.8	27.62 [3.49, 218.57]	27.54 [3.17, 239.45]	5.07 [1.27, 20.31]	5.43 [1.14, 26.01]	1.95 [0.5, 7.55]	1.85 [0.41, 8.25]
Emotional intimate partner violence^c
None (0)	360	90.7	Reference	Reference	Reference	Reference	Reference	Reference
Some (1)	21	5.3	4.53 [1.82, 11.28]	4.41 [1.62, 11.97]	1.24 [0.28, 5.60]	1.15 [0.24, 5.53]	1.13 [0.37, 3.48]	1.07 [0.32, 3.58]
High (≥2)	16	4.0	8.37 [2.64, 26.58]	14.20 [3.91, 51.49]	3.93 [1.19, 12.99]	4.83 [1.12, 20.76]	1.60 [0.5, 5.13]	2.01 [0.54, 7.50]

Note. PTSD = posttraumatic stress disorder; OR = odds ratio; CI = confidence interval.

Adjusted for age, race, education, marital status, and smoking.

Sum of “yes” responses for physical and emotional intimate partner violence.

Sum of three questions, range 0-3.

Next, unadjusted and multivariable (adjusted) logistic regression was utilized to investigate the association between indicators of neighborhood crime and safety and mental health outcomes (Table 3). Women who reported medium and high neighborhood safety had significantly higher odds of probable depression after adjustment for important confounders than those who had perceived low levels of neighborhood safety. Women who were not sure if they would or would not report a crime if they witnessed it were more likely to have depression than those who would report a crime to police. Similarly, medium and high total neighborhood violence were significantly associated with increased odds of probable depression, when compared with low neighborhood violence. Perceived neighborhood safety was not significantly associated with PTSD although women who were not sure if they would report a crime if witnessed or would not report a crime were more likely to have PTSD than those who would report. High total neighborhood violence was also significantly associated with increased odds of PTSD. After adjustment, none of the neighborhood crime indicators were associated with PA.

Table 3.

Unadjusted and Adjusted^a Multivariable Logistic Regression of Neighborhood Violence Indicators and Mental Health Outcomes (N = 398).

			Probable Depression		PTSD		Pregnancy-Specific Anxiety
			Unadjusted	Adjusted	Unadjusted	Adjusted	Unadjusted	Adjusted
	N	%	OR [95% CI]	OR [95% CI]	OR [95% CI]	OR [95% CI]	OR [95% CI]	OR ]95% CI]
Perceived neighborhood safety
Low	179	45.0	Reference	Reference	Reference	Reference	Reference	Reference
Medium	23	23.4	2.24 [1.27, 3.95]	1.88 [1.02, 3.48]	1.49 [0.50, 4.43]	1.10 [0.36, 3.40]	1.30 [0.65, 2.64]	1.05 [0.50, 2.19]]
High	126	31.7	3.30 [1.98, 5.52]	2.11 [1.19, 3.72]	4.07 [1.73, 9.58]	2.32 [0.92, 5.85]	2.31 [1.28, 4.18]	1.61 [0.84, 3.08]
Would report a crime if witnessed
Yes	265	66.9	Reference	Reference	Reference	Reference	Reference	Reference
Not sure	104	26.3	2.46 [1.51, 4.00]	1.80 [1.05, 3.07]	4.17 [1.86, 9.32]	3.47 [1.48, 8.13]	1.39 [0.78, 2.46]	1.12 [0.60, 2.09]
No	27	6.8	6.98 [2.98, 16.35]	3.53 [1.4, 8.89]	8.02 [2.8, 22.94]	4.82 [1.53, 15.15]	1.17 [0.42, 3.27]	0.64 [0.21, 1.92]
Total neighborhood violence
Low	130	32.7	Reference	Reference	Reference	Reference	Reference	Reference
Medium	177	44.5	4.25 [2.25, 8.03]	2.13 [1.20, 3.77]	3.12 [0.86, 11.28]	2.98 [0.96, 9.23]	1.63 [0.86, 3.11]	1.35 [0.71, 2.60]
High	91	22.9	8.85 [4.44, 17.65]	2.70 [1.45, 5.04]	11.17 [3.20, 39.05]	6.54 [2.18, 19.57]	2.14 [1.05, 4.37]	1.12 [0.54, 2.32]

Note. PTSD = posttraumatic stress disorder; OR = odds ratio; CI = confidence interval.

Adjusted for age, race, education, marital status and smoking.

Finally, unadjusted and multivariable (adjusted) logistic regression was used to investigate whether cumulative (intimate partner and neighborhood) violence was associated with mental health outcomes (Table 4). Findings are reported for both equal weighting (both types of violence weighted equally), and for weighted cumulative violence indicators (types of violence weighted by association with outcome). After adjustment, women with medium (middle tertile) total violence and those with high total violence (top tertile) had increased odds of having probable depression when compared with those with low violence (in the lowest tertile; equal weighting). Women with medium and high total violence (weighted) had higher odds of probable depression than those with low total violence. Findings for PTSD demonstrated that both high total violence (equal weighting) and high total violence (weighted) were associated with increased odds of the disorder, while there was not a significant increased odds in medium total violence (equal weighting or weighted) compared with low groups. Adjusted models did not reveal a statistically significant association between total violence measures and PA.

Table 4.

Unadjusted and Adjusted^a Multivariable Regression of Cumulative Violence Indicator (Total Violence) and Mental Health Outcomes (N = 398).

	Probable Depression		PTSD		Pregnancy-Specific Anxiety
	Unadjusted	Adjusted	Unadjusted	Adjusted	Unadjusted	Adjusted
	OR [95% CI]	OR [95% CI]	OR [95% CI)]	OR [95% CI]	OR [95% CI]	OR [95% CI]
Total violence, equal weighting
Low	Reference		Reference	Reference	Reference	Reference
Medium	3.39 [1.88, 6.13]	2.45 [1.29, 4.65]	2.17 [0.67, 7.08]	1.45 [0.42, 5.04]	1.60 [0.80, 3.17]	1.21 [0.58, 2.53]
High	5.53 [3.29, 9.31]	3.73 [2.13, 6.55]	7.48 [3.08, 18.15]	5.45 [2.12, 14.02]	1.81 [1.01, 3.27]	1.28 [0.67, 2.44]
Total violence, weighted
Low	Reference	Reference	Reference	Reference	Reference	Reference
Medium	2.78 [1.49, 5.20]	1.88 [0.94, 3.73]	2.08 [0.64, 6.77]	1.44 [0.42, 4.99]	1.60 [0.8, 3.17]	1.21 [0.58, 2.53]
High	5.85 [3.5, 9.75]	4.14 [2.39, 7.19]	7.75 [3.19, 18.81]	5.54 [2.15, 14.29]	1.81 [1.01, 3.27]	1.28 [0.67, 2.44]

Note. PTSD = posttraumatic stress disorder; OR = odds ratio; CI = confidence interval.

Adjusted for age, race, education, marital status, and smoking.

As IPV and neighborhood violence are predictive of the outcomes separately, the question arises whether a combined measure is a better predictor than the individual measures. One possibility is that different types of violence are correlated with each other, so that the relationship observed both represent the effect of one confounded by the other. We did not find this to be likely, as both kinds of violence were significant when included in a single model. Next, we examined whether a cumulative measure was more predictive of the outcome than individual measures using logistic regression to examine the area under the receiver operating curve (aROC). The predictive power of the total violence measure was significantly better than the neighborhood violence measure, and was higher but not statistically stronger than the predictive power of the IPV model. IPV was strongly associated with poorer mental health, but the CI was very wide due to the small number of exposed. Therefore, the cumulative measure had better precision, but the predictive power of the model was not statistically stronger.

Discussion

In this cross-sectional study of women living in metropolitan New Orleans between 5 and 7 years after Hurricane Katrina, we found, after adjustment for important risk factors, that intimate partner (physical, emotional, and overall) violence was associated with higher odds of probable depression and PTSD and that physical domestic violence alone was associated with high odds of PA. Indicators of neighborhood crime and safety were also significantly associated with probable depression and PTSD. When both intimate partner violence and neighborhood crime were combined, cumulative violence (equal weighting and weighted) was associated with increased odds of probable depression and PTSD, but not PA.

The observed prevalence of probable depression was 30.4% in our study, which is slightly higher than the observed EDS in two cohort studies of primiparas in the Northeast. In both of those studies, women were interviewed earlier in pregnancy (before 22 weeks gestation), compared with our study (mean gestational age at interview = 31 weeks). Among African American pregnant women enrolled in Project Viva (N = 113), 14.2% had EDS scores >12 (Ertel et al., 2012). Depression was also measured in a prospective cohort study (conducted in 2006-2007) of demographically similar New Orleans women who were pregnant during Hurricane Katrina or soon afterwards (N = 199). In that sample, 11.6% of women had probable depression (EDS > 12; Xiong et al., 2008). We observed that 17.6% of participants had PA, while others have reported that 21% to 24% of women have anxiety disorders in pregnancy (Grant, McMahon, & Austin, 2008; Sutter-Dallay, Giaconne-Marcesche, Glatigny-Dallay, & Verdoux, 2004). The prevalence of PTSD in the current study was 9.2% (mean = 29.6). A study of New Orleans women post-Katrina reported 23.1% had a PCL >50 (Xiong et al., 2008). A longitudinal study (N = 119) in Hawaii assessed PTSD four times during pregnancy and reported a mean PCL of 27.4 between 28.0 and 32.9 weeks gestation (Onoye et al., 2013).

Our findings were consistent with previous studies (Dutton, Kaltman, Goodman, Weinfurt, & Vankos, 2005; Hathaway et al., 2000; Pico-Alfonso et al., 2006; Seng, Low, Sperlich, Ronis, & Liberzon, 2009) linking intimate partner violence with adverse mental health outcomes. A secondary analysis (N = 182) found that exposure to a combination of physical and psychological IPV had higher rates of depression, anxiety, and PTSD symptoms (Pico-Alfonso et al., 2006). A population-based study (N = 2,043) reported that women who disclosed IPV had significantly increased risk of anxiety (Relative Risk [RR] = 2.2, 95% CI = [1.1, 4.4) than those not reporting abuse (Hathaway et al., 2000). Another sample of 1,581 nulliparous women (<28 weeks gestation) reported that women who experienced IPV had significantly higher odds of PTSD (OR 11.9, 95% CI = [3.6, 39.9]; Seng et al., 2009). Finally, another study (N = 406) reported that women who experienced physical and psychological abuse but little sexual violence were more likely to have major depressive disorders (OR 2.50, 95% CI = [1.22, 5.12]; Dutton et al., 2005) and those who also reported sexual violence were more likely to have symptoms of PTSD than those with physical, psychological abuse but little sexual violence (OR 2.51, 95% CI = [1.43, 4.41]).

Similarly, our finding that exposure to neighborhood violence is associated with poorer mental health in pregnancy reinforces some previous work (Casciano & Massey, 2012; Giurgescu et al., 2015), but not all (Yang, Kestens, Dahhou, Daniel, & Kramer, 2015). A prospective study of 95 pregnant African American women and found that depressive symptoms were associated with perceived measures of neighborhood violence, including physical disorder (r = .380, p < .01), social disorder (r = .324, p < .01), and crime (r = .280, p < .01; Giurgescu et al., 2015). A cross-sectional study examined the association between a neighborhood disorder scale and anxiety symptoms among African Americans (N = 116) and reported that residents living in projects were exposed to less neighborhood violence than nonresidents and experienced fewer anxiety symptoms than nonresidents (1.76 vs. 2.06 [on a scale of 1-5], p = .01; Casciano & Massey, 2012). In contrast, Yang et al. (2015) studied 5,337 pregnant women (24-26 weeks) and reported a null association between high neighborhood material (β = .65, 95% CI = [–0.01, 1.32]) and social deprivation and psychological distress (β = .07, 95% CI = [–0.56, 0.72]) after adjustment (Yang et al., 2015). The differences in reported effect sizes and findings may be explained by inconsistent definitions and measurement of neighborhood itself, as well as disorder and crime.

Finally, the few studies that investigated cumulative effects of IPV and neighborhood violence on mental health have demonstrated mixed results; two reported similar positive findings (Brown et al., 2005; Clark et al., 2008) and one did not (Bogat et al., 2005). We found that cumulative violence was associated with 7 to 14 times increased odds of PTSD and probable depression. Differences in findings here were likely due to variation in measurement of neighborhood violence, as perception of community crime and violence is more likely to be associated with women’s mental health than objective measures (Giurgescu et al., 2015).

This study offers additional insight on how violence at multiple levels and neighborhood disorder contribute to individual health and illness, and has both strengths and limitations. Strengths include combined focus on individual experiences of violence as well as perception of neighborhood factors. One limitation of the study is the relatively small sample size. This was particularly the case in the highly exposed IPV group and for PTSD, and led to unstable estimates for some ORs. In addition, social desirability bias may be present, as women were interviewed by apparent strangers regarding crime and safety in their neighborhoods, and may have felt pressure to respond to questions in a socially acceptable way. We found that women who would not report a crime if they witnessed it were more likely to have symptoms of mental illness, which may indicate feelings of stress related to these crimes and could have led to underreporting and possible residual confounding. Another limitation is that we did not collect information on eligible women who declined to participate in the study. Although we studied a convenience sample of pregnant women, it may be that women with worse mental health declined to participate, which would have biased our results toward the null.

Future research should include a broader description and understanding of exposure to multiple types of violence over the life course, including cumulative effects on women, and how diversity of socioeconomic status and ethnic/racial backgrounds affect disaster recovery. Consideration of larger social issues, such as political and social disempowerment, may improve our understanding of the mechanisms at play. Women who mistrust the criminal justice system or fear for their own personal safety are not likely to report crimes, and improved political empowerment of African Americans may be part of the answer to improve perinatal health in disadvantaged communities (Bell, Zimmerman, Almgren, Mayer, & Huebner, 2006).

This research adds to the limited body of knowledge around cumulative effects of violence on pregnant women, and has implications for improved clinical practice. Violence occurs commonly in pregnancy, and all women should be screened for IPV during prenatal care. We would like to propose further recommendations that prenatal care providers in communities affected by violence (which are also often affected by disaster) devote increased attention to IPV screening. Providers can ascertain during prenatal visits if there is a gun in the home as part of prenatal care and health promotion for homes with young children. Further assessment of women who screen positive for IPV should then include a lethality assessment, safety planning, and referral to counseling services. Community-based interventions are also necessary, and could include collaboration with law enforcement to strengthen policies for removal of guns from homes where there is a known history of IPV. Finally, high-quality programs for perpetrators of abuse (Campbell, Glass, Sharps, Laughon, & Bloom, 2007) must not be forgotten as an important intervention to support the health expectant mothers and families.

Data from this study are available upon written request to the authors.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This article was supported by National Institutes of Health (NIH), National Institute of Nursing Research (NINR), 5R03NR012052-02, and National Institute of Child Health and Human Development (NICHD) T32HD057780.

Author Biographies

Veronica Barcelona de Mendoza, PhD, MSN, RN, APHN-BC, is an advanced practice public health nurse with a doctorate in epidemiology. She has worked in local and international maternal and child health for several years and is currently a postdoctoral associate at Yale School of Nursing.

Emily W. Harville, PhD, has a doctorate in epidemiology and has worked for several years in the area of perinatal and reproductive health. She is an expert on perinatal health and disasters, having conducted several studies and published extensively on the health of women and children post-Hurricane Katrina.

Jane Savage, PhD, MS, RN, CNE, LCCE, FACCE, is a seasoned nurse and educator who has taught mental health nursing and maternal and child health nursing for more than 30 years. She is has been involved in disaster research since 2005.

Gloria Giarratano, PhD, APRN, CNS, is an advanced practice nurse who has taught maternal/child nursing for many years and has been involved in disaster research and preparedness in her local community.

References

Adewuya

A. O.

Ola

B. A.

Dada

A. O.

Fasoto

O. O.

(2006). Validation of the Edinburgh Postnatal Depression Scale as a screening tool for depression in late pregnancy among Nigerian women. Journal of Psychosomatic Obstetrics & Gynecology, 27, 267-272.

Adouard

Glangeaud-Freudenthal

N. M. C.

Golse

(2005). Validation of the Edinburgh Postnatal Depression Scale (EPDS) in a sample of women with high-risk pregnancies in France. Archives of Women’s Mental Health, 8, 89-95.

Aisenberg

Herrenkohl

(2008). Community violence in context: Risk and resilience in children and families. Journal of Interpersonal Violence, 23, 296-315. doi:10.1177/0886260507312287

Ajrouch

K. J.

Reisine

Lim

Sohn

Ismail

(2010). Situational stressors among African American women living in low-income urban areas: The role of social support. Women & Health, 50, 159-175. doi:10.1080/03630241003705045

Alderdice

Lynn

Lobel

(2012). A review and psychometric evaluation of pregnancy-specific stress measures. Journal of Psychosomatic Obstetrics and Gynecology, 33, 62-77.

Alvarado-Esquivel

Sifuentes-Alvarez

Salas-Martinez

Martinez-Garcia

(2006). Validation of the Edinburgh postpartum depression scale in a population of puerperal women in Mexico. Clinical Practice in Epidemiological Mental Health, 29, 2-33.

Amaro

Vega

R. R.

Valencia

(2001). Gender, context and HIV prevention among Latinos. In Aguirre-Molina

Molina

Zambrana

R. E.

(Eds.), Health issues in the Latino community (1st ed., pp. 301-326). San Francisco, CA: Jossey-Bass.

Andrykowski

M. A.

Cordova

M. J.

Studts

J. L.

Miller

T. W.

(1998). Posttraumatic stress disorder after treatment for breast cancer: Prevalence of diagnosis and use of the PTSD checklist-civilian version (PCL-C) as a screening instrument. Journal of Consulting and Clinical Psychology, 66, 586-590.

Bell

J. F.

Zimmerman

F. J.

Almgren

G. R.

Mayer

J. D.

Huebner

C. E.

(2006). Birth outcomes among urban African American women: A multilevel analysis of the role of racial residential segregation. Social Science & Medicine, 63, 3030-3045.

10.

Bogat

G. A.

Leahy

von Eye

Maxwell

Levendosky

A. A.

Davidson

W. S.

(2005). The influence of community violence on the functioning of women experiencing domestic violence. American Journal of Community Psychology, 36, 123-132. doi:10.1007/s10464-005-6237-2

11.

Brown

J. R.

Hill

H. M.

Lambert

S. F.

(2005). Traumatic stress symptoms in women exposed to community and partner violence. Journal of Interpersonal Violence, 20, 1478-1494.

12.

Campbell

J. C.

Glass

Sharps

P. W.

Laughon

Bloom

(2007). Intimate partner homicide: Review and implications of research and policy. Trauma, Violence, & Abuse, 8, 246-269.

13.

Casciano

Massey

D. S.

(2012). Neighborhood disorder and anxiety symptoms: New evidence from a quasi-experimental study. Health & Place, 18, 180-190. doi:10.1016/j.healthplace.2011.09.002

14.

Centers for Disease Control and Prevention, Division of Reproductive Health at National Center for Chronic Disease Prevention and Health Promotion. (2009). Pregnancy Risk Assessment Monitoring System (PRAMS). Retrieved from http://www.cdc.gov/prams/questionnaire.htm

15.

Chang

Berg

C. J.

Saltzman

L. E.

Herndon

(2005). Homicide: A leading cause of injury deaths among pregnant and postpartum women in the United States, 1991-1999. American Journal of Public Health, 95, 471-477.

16.

Clark

Ryan

Kawachi

Canner

M. J.

Berkman

Wright

R. J.

(2008). Witnessing community violence in residential neighborhoods: A mental health hazard for urban women. Journal of Urban Health: Bulletin of the New York Academy of Medicine, 85, 22-38. doi:10.1007/s11524-007-9229-8

17.

Clements

A. D.

Bailey

B. A.

(2010). The relationship between temperament and anxiety: Phase I in the development of a risk screening model to predict stress-related health problems. Maternal Child Health Journal, 17, 1151-1157.

18.

CommonHealth ACTION. (2009). New Orleans: Community perspectives on the root causes of violence. Washington, DC: CommonHealth ACTION National Office.

19.

Davidson

T. M.

Price

McCauley

J. L.

Ruggiero

K. J.

(2013). Disaster impact across cultural groups: Comparison of Whites, African Americans, and Latinos. American Journal of Community Psychology, 52, 97-105. doi:10.1007/s10464-013-9579-1

20.

Dutton

M. A.

Kaltman

Goodman

L. A.

Weinfurt

Vankos

(2005). Patterns of intimate partner violence: Correlates and outcomes. Violence and Victims, 20, 483-497.

21.

Earls

F. J.

Brooks-Gunn

Raudenbush

Sampson

(1994). The project on human development in Chicago neighborhoods, community survey. U.S. Department of Justice. Retrieved from http://www.icpsr.umich.edu/icpsrweb/PHDCN/studies/02766

22.

Egan

Tannahill

Petticrew

Thomas

(2008). Psychosocial risk factors in home and community settings and their associations with population health and health inequalities: A systematic meta-review. BMC Public Health, 8, 239. doi:10.1186/1471-2458-8-239

23.

El Kady

Gilbert

W. M.

Xing

Smith

L. H.

(2005). Maternal and neonatal outcomes of assaults during pregnancy. Obstetrics & Gynecology, 105, 357-363.

24.

Ertel

K. A.

James-Todd

Kleinman

Krieger

Gillman

Wright

Rich-Edwards

(2012). Racial discrimination, response to unfair treatment, and depressive symptoms among pregnant Black and African American women in the United States. Annals of Epidemiology, 22, 840-846. doi:10.1016/j.annepidem.2012.10.001

25.

Federal Bureau of Investigation. (2012). Crime in the United States, 2012: Uniform crime reports. Washington, DC: U.S. Department of Justice.

26.

Fothergill

Maestas

E. G.

Darlington

J. D.

(1999). Race, ethnicity and disasters in the United States: A review of the literature. Disasters, 23, 156-173.

27.

Gillespie

C. F.

Bradley

Mercer

Smith

A. K.

Conneely

Gapen

. . . Ressler

K. J.

(2009). Trauma exposure and stress-related disorders in inner city primary care patients. General Hospital Psychiatry, 31, 505-514. doi:10.1016/j.genhosppsych.2009.05.003

28.

Giurgescu

Zenk

S. N.

Templin

T. N.

Engeland

C. G.

Dancy

B. L.

Park

C. G.

. . . Misra

D. P.

(2015). The impact of neighborhood environment, social support, and avoidance coping on depressive symptoms of pregnant African American women. Women’s Health Issues: Official Publication of the Jacobs Institute of Women’s Health, 25, 294-302. doi: 10.1016/j.whi.2015.02.001

29.

Goldmann

Aiello

Uddin

Delva

Koenen

Gant

L. M.

Galea

(2011). Pervasive exposure to violence and posttraumatic stress disorder in a predominantly African American urban community: The Detroit neighborhood health study. Journal of Traumatic Stress, 24, 747-751. doi:10.1002/jts.20705

30.

Grant

K. A.

McMahon

Austin

M. P.

(2008). Maternal anxiety during the transition to parenthood: A prospective study. Journal of Affective Disorders, 108, 101-111.

31.

Harville

E. W.

Taylor

C. A.

Tesfai

Buekens

(2011). Experience of hurricane Katrina and reported intimate partner violence. Journal of Interpersonal Violence, 26, 833-845. doi:10.1177/0886260510365861

32.

Hathaway

J. E.

Mucci

L. A.

Silverman

J. G.

Brooks

D. R.

Mathews

Pavlos

C. A.

(2000). Health status and health care use of Massachusetts women reporting partner abuse. American Journal of Preventive Medicine, 19, 302-307. doi:S0749-3797(00)00236-1

33.

James

Brody

Hamilton

(2013). Risk factors for domestic violence during pregnancy: A meta-analytic review. Violence and Victims, 28, 359-380.

34.

Kaiser Family Foundation. (2007). Giving voice to the people of New Orleans: The Kaiser post-Katrina baseline survey. Menlo Park, CA: Author.

35.

Kelly

Hall

(2010). Measuring anxiety in adolescents exposed to community violence: A review, comparison, and analysis of three measures. Issues in Mental Health Nursing, 31, 28-38. doi:10.3109/01612840903200027

36.

Lagdon

Armour

Stringer

(2014). Adult experience of mental health outcomes as a result of intimate partner violence victimisation: A systematic review. European Journal of Psychotraumatology, 5. doi:10.3402/ejpt.v5.24794

37.

Miles

J. N. V.

Marshall

G. N.

Schell

T. L.

(2008). Spanish and english versions of the PTSD checklist – civilian version (PCL-C): Testing for differential item functioning. Journal of Traumatic Stress 21(4), 369-376. doi:10.1002/jts.20349

38.

Matthey

Henshaw

Elliott

Barnett

(2006). Variability in use of cut-off scores and formats on the Edinburgh postnatal depression scale: Implications for clinical and research practice. Archives of Women’s Mental Health, 9, 309-315. doi:10.1007/s00737-006-0152-x

39.

McFarlane

Campbell

J. C.

Sharps

Watson

(2002). Abuse during pregnancy and femicide: Urgent implications for women’s health. Obstetrics & Gynecology, 100, 27-36.

40.

Messer

L. C.

Maxson

Miranda

M. L.

(2013). The urban built environment and associations with women’s psychosocial health. Journal of Urban Health: Bulletin of the New York Academy of Medicine, 90, 857-871. doi:10.1007/s11524-012-9743-1

41.

Onoye

J. M.

Shafer

L. A.

Goebert

D. A.

Morland

L. A.

Matsu

C. R.

Hamagami

(2013). Changes in PTSD symptomatology and mental health during pregnancy and postpartum. Archives of Women’s Mental Health, 16, 453-463. doi:10.1007/s00737-013-0365-8

42.

Phillips

B. D.

Thomas

D. S. K.

Fothergill

Blinn-Pike

(Eds.). (2010). Social vulnerability in disasters (1st ed.). Boca Raton, FL: CRC Press.

43.

Pico-Alfonso

M. A.

Garcia-Linares

M. I.

Celda-Navarro

Blasco-Ros

Echeburua

Martinez

(2006). The impact of physical, psychological, and sexual intimate male partner violence on women’s mental health: Depressive symptoms, posttraumatic stress disorder, state anxiety, and suicide. Journal of Women’s Health, 15, 599-611. doi:10.1089/jwh.2006.15.599

44.

Pinchevsky

G. M.

Wright

E. M.

(2012). The impact of neighborhoods on intimate partner violence and victimization. Trauma, Violence, & Abuse, 13, 112-132. doi:10.1177/1524838012445641

45.

Seng

J. S.

Low

L. K.

Sperlich

Ronis

D. L.

Liberzon

(2009). Prevalence, trauma history, and risk for posttraumatic stress disorder among nulliparous women in maternity care. Obstetrics & Gynecology, 114, 839-847. doi:10.1097/AOG.0b013e3181b8f8a2

46.

Shah

P. S.

Shah

, Knowledge Synthesis Group on Determinants of Preterm/LBW Births. (2010). Maternal exposure to domestic violence and pregnancy and birth outcomes: A systematic review and meta-analyses. Journal of Women’s Health, 19, 2017-2031. doi:10.1089/jwh.2010.2051

47.

Silverman

J. G.

Decker

M. R.

Reed

Raj

(2006). Intimate partner violence around the time of pregnancy: Association with breastfeeding behavior. Journal of Women’s Health, 15, 934-940. doi:10.1089/jwh.2006.15.934

48.

SOSNA/NAC/CDC Public Safety Committee. (n.d.). Resident survey on police services. Retrieved from http://www.southphillyblocks.org/sosna/site_archive/about/police.html

49.

Sutter-Dallay

A. L.

Giaconne-Marcesche

Glatigny-Dallay

Verdoux

(2004). Women with anxiety disorders during pregnancy are at increased risk of intense postnatal depressive symptoms: A prospective survey of the MATQUID cohort. European Psychiatry: The Journal of the Association of European Psychiatrists, 19, 459-463.

50.

Taylor

S. E.

Holden

K. B.

(2009). The health status of Black women. In Braithwaite

R. L.

Taylor

S. E.

Treadwell

H. M.

(Eds.), Health issues in the Black community (3rd ed., pp. 53-72). San Francisco, CA: Jossey-Bass.

51.

Tracy

Morgenstern

Zivin

Aiello

A. E.

Galea

(2014). Traumatic event exposure and depression severity over time: Results from a prospective cohort study in an urban area. Social Psychiatry and Psychiatric Epidemiology, 49, 1769-1782. doi:10.1007/s00127-014-0884-2

52.

Urquia

M. L.

O’Campo

P. J.

Heaman

M. I.

Janssen

P. A.

Thiessen

K. R.

(2011). Experiences of violence before and during pregnancy and adverse pregnancy outcomes: An analysis of the Canadian maternity experiences survey. BMC Pregnancy & Childbirth, 11, 42. doi:10.1186/1471-2393-11-42

53.

Ventureyra

V. A.

Yao

S. N.

Cottraux

Note

De Mey-Guillard

(2002). The validation of the posttraumatic stress disorder checklist scale in posttraumatic stress disorder and nonclinical subjects. Psychotherapy and Psychosomatics, 71, 47-53.

54.

Wilson-Genderson

Pruchno

(2013). Effects of neighborhood violence and perceptions of neighborhood safety on depressive symptoms of older adults. Social Science & Medicine, 85, 43-49. doi:10.1016/j.socscimed.2013.02.028

55.

Xiong

Harville

E. W.

Mattison

D. R.

Elkind-Hirsch

Pridjian

Buekens

(2008). Exposure to hurricane Katrina, post-traumatic stress disorder and birth outcomes. The American Journal of the Medical Sciences, 336, 111-115. doi:10.1097/MAJ.0b013e318180f21c

56.

Yali

A. M.

Lobel

(1999). Coping and distress in pregnancy: An investigation of medically high risk women. Journal Psychosomatic Obstetrics & Gynaecology, 20, 39-52.

57.

Yang

Kestens

Dahhou

Daniel

Kramer

M. S.

(2015). Neighborhood deprivation and maternal psychological distress during pregnancy: A multilevel analysis. Maternal and Child Health Journal, 19, 1142-1151. doi:10.1007/s10995-014-1623-8