Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Apr 1.
Published in final edited form as: Matern Child Health J. 2020 Nov 18;25(4):635–644. doi: 10.1007/s10995-020-03026-x

Associations of maternal trait anger expression and lifetime traumatic and non-traumatic experiences with preterm birth

Whitney Cowell 1, Lilly Taing 2, Talia Askowitz 3, Michelle Bosquet Enlow 4,5, Michele R Hacker 6,7,8, Rosalind J Wright 1,3
PMCID: PMC8035151  NIHMSID: NIHMS1647938  PMID: 33206302

Abstract

Objectives:

Most studies examining psychosocial factors contributing to preterm birth (PTB) have focused on negative life events. Studies examining the influence of negative emotion, in particular maternal anger, remain sparse. We examined associations of maternal trait anger expression and lifetime traumatic and non-traumatic experiences with the risk of PTB.

Methods:

Mother-newborn pairs were enrolled in the PRogramming of Intergenerational Stress Mechanisms pregnancy cohort based in Boston and New York City. Women completed the State-Trait Anger Expression Inventory-2 (STAXI-2), Life Stressor Checklist-Revised (LSC-R), and Childhood Trauma Questionnaire (CTQ) in pregnancy. We used modified Poisson regression to estimate the relative risk (RR) of PTB 1) in relation to continuous STAXI-2 trait Anger Expression-In (AX-I) and trait Anger Expression-Out (AX-O) subscales, 2) in relation to continuous LSC-R scores, and 3) between women who did versus did not experience childhood sexual, emotional, and/or physical abuse in six separate models. We also examined interactions between maternal anger expression and lifetime stress/childhood trauma.

Results:

Younger, single, minority women had higher outward anger expression and inward anger suppression. Trait AX-I and AX-O scores were higher among women who experienced abuse in childhood and who had higher lifetime stress. Maternal lifetime stress, outward anger expression, and inward anger suppression were associated with an increased risk of PTB in separate models; however, stress, trauma and anger did not interact to further increase the risk of PTB.

Conclusions:

Higher anger expression and higher lifetime stress experiences were associated with an increased risk of PTB among a racially and ethnically diverse sample of pregnant women.

Significance:

Preterm birth (PTB) is a leading cause of infant mortality and morbidity and has consequences for long-term health. Accumulating evidence supports that psychosocial stress and stress correlates (e.g., depression) contribute to the pathophysiology underlying adverse obstetrical outcomes, including PTB. Few studies have examined the influence of negative emotions, including anger expression, on adverse birth outcomes. In this study, higher outward anger expression and inward anger suppression were associated with an increased risk of PTB even after adjusting for potential confounders. Lifetime stress was also positively associated with PTB, but childhood abuse subscales were not.

Keywords: stress, trauma, anger, pregnancy, preterm birth

Introduction

The United States ranks 32nd out of the 35 wealthiest nations in the incidence of infant mortality (OECD, 2018). Approximately 10% of infants are born preterm (<37 weeks of gestation), and disorders related to short gestation and low birthweight, a correlate of preterm birth (PTB), are the second leading cause of death among infants under one year (Martin, Hamilton, Osterman, Driscoll, & Drake, 2018). In addition to neonatal morbidities, the consequences of PTB persist across the life course and include an increased risk of neurodevelopmental disorders, metabolic syndrome, and other chronic diseases, with substantial individual and societal burdens (Moster, Lie, & Markestad, 2008).

Despite the high incidence of PTB, a large proportion of preterm deliveries are precipitated by unknown risk factors (Muglia & Katz, 2010). Accumulating evidence supports that a range of psychosocial factors, including traumatic and non-traumatic life stressors experienced across a woman’s life course, contribute to the pathophysiology underlying adverse obstetrical and infant outcomes, including an increased risk of PTB (Muglia & Katz, 2010). Studies linking adverse childhood experiences, including abuse, with an increased risk of PTB suggest that traumatic experiences extending from one’s own childhood may be particularly consequential (Christiaens, Hegadoren, & Olson, 2015). Other research has linked negative mood states, including depression and anxiety, in pregnant women with adverse birth outcomes, including PTB (Staneva, Bogossian, Pritchard, & Wittkowski, 2015). Fewer studies have examined the influence of negative emotions on adverse birth outcomes.

How one manifests and expresses anger has been linked with several psychopathologies and chronic health conditions in both non-pregnant and pregnant samples. Outwardly negative and poorly controlled expressions of anger are associated with depression, aggression, and worse self-reported health, whereas the suppression of angry feelings has been linked to cardiovascular morbidity and generalized anxiety disorder (Deschenes, Dugas, Fracalanza, & Koerner, 2012; Suls, 2013). Both outward and inward expressions of anger are associated with an increased risk for hypertension, a higher prevalence of metabolic syndrome, and worse coronary heart disease prognosis among women (Laszlo, Janszky, & Ahnve, 2010). Findings from limited research have also linked anger during pregnancy with changes in fetal neurobiology and developmental trajectories (Field et al., 2002; Wei et al., 2017), suggesting that maternal anger may impact the developing fetus. Yet, no prior study has examined the association between anger expression in pregnancy and adverse birth outcomes. We address this research gap by investigating associations between maternal trait anger expression and infant PTB in an urban, multiracial pregnancy cohort. We hypothesized that higher maternal anger expression would be associated with an increased risk for infant PTB.

We also examined associations with maternal stress and trauma and explored interactions between trait anger expression and stress/trauma on the risk of PTB. Similar to stress experienced over the life course, trait affect has been associated with hypothalamic-pituitary-adrenal (HPA) axis activity, and research has shown that exposure to chronic stress and accompanying chronic negative affective arousal is linked to persistent alterations in cortisol regulation (Polk, Cohen, Doyle, Skoner, & Kirschbaum, 2005). For example, trait anger has been associated with modified HPA axis activity as indicated by altered features of the diurnal cortisol rhythm, with similar patterns observed in response to chronic strain and perceived stress (Adam, 2006). Childhood may be an especially vulnerable period, with research supporting that exposure to early life stress leads to chronic alterations in HPA axis functioning (Karlamangla et al., 2019). Additionally, greater emotional reactivity and expression have been associated with poorer social interactions; thus, outward anger expression may increase vulnerability to psychosocial stress via fewer social supports and social alienation. Based on this evidence, we hypothesized that maternal anger and lifetime experiences of stress and trauma would interact to further increase the risk of PTB.

Methods

Study sample

Participants included mother-newborn pairs enrolled in the PRogramming of Intergenerational Stress Mechanisms (PRISM) study, a prospective pregnancy cohort designed to investigate the long-term health consequences of perinatal stress exposure on child development. Pregnant women were recruited from prenatal care clinics and hospital obstetrical units in Boston and New York City beginning in 2011. Eligibility criteria for the PRISM cohort included (1) English or Spanish-speaking, and (2) age ≥ 18 years at enrollment. Exclusion criteria for the PRISM cohort included (1) multiple gestation pregnancy, (2) maternal intake of ≥ 7 alcoholic drinks per week before pregnancy recognition or any alcohol after pregnancy recognition, (3) maternal HIV positive status, or (4) pregnancy loss or the presence of a major congenital or genetic disorder identified during pregnancy or at birth. At the time of this study, 870 eligible women were enrolled and had delivered a live born infant without a major health or genetic concern. We restricted the analytic sample to 829 (95%) women who self-reported White, Black, or Hispanic race/ethnicity; we excluded the 5% of women reporting other race/ethnicity due to the small number of preterm births. Complete exposure (anger, stress, trauma; see Figure 1) and outcome (PTB) data were available for 658 of these 829 women (80%). We excluded an additional 30 participants (4.5%) with missing covariate data, resulting in an analytic sample of 628 mother-newborn pairs. Written informed consent was obtained from women prior to study participation in their preferred language. All study procedures were approved by the institutional review boards at the Brigham and Women’s Hospital and the Icahn School of Medicine at Mount Sinai; Beth Israel Deaconess Medical Center relied on Brigham and Women’s Hospital for review and oversight of the protocol.

Figure 1.

Figure 1.

Open in a new tab

Overview of the content and timing of study assessments administered.

Maternal trait anger expression

Women completed the trait anger expression subscales on the State-Trait Anger Expression Inventory-2 (STAXI-2) during an in-person interview conducted between 24–42 weeks gestation. The STAXI-2 has high face validity, concurrent validity with other anger assessment tools, construct validity, internal reliability, and test-retest reliability (Spielberger, 1999). The STAXI-2 is validated for use in English and Spanish and is one of the most commonly used research instruments for assessing how anger contributes to mental and physical illnesses. We administered the trait anger expression scales, which ask women to indicate how often they generally react or behave when they are feeling angry as described by a series of items. The Anger Expression-Out (AX-O) subscale measures the degree to which angry feelings are expressed outwardly and includes items such as “when I feel angry I do things like slam doors.” The Anger Expression-In (AX-I) subscale measures suppression of angry feelings and includes items such as “when I feel angry I boil inside but don’t show it.” Each subscale comprises eight items, which are rated on a four-point Likert scale as: 1-never true, 2-sometimes true, 3-often true, or 4-almost always true. We summed the items in each subscale to create two summary measures, each with a possible range from 8–32.

Lifetime traumatic and non-traumatic stress experiences

Participants completed the Life Stressor Checklist-Revised (LSC-R) scale during an in-person interview conducted between 24–27 weeks gestation. The 30-item LSC-R provides a measure of lifetime exposure to stressful and potentially traumatic events (e.g., accident or natural disasters, violence, incarceration) (Wolfe, Kimerling, Brown, Chrestman, & Levin, 1996). The LSC-R has established test-retest reliability and validity in diverse populations and is particularly well-suited for women, as it includes questions related to interpersonal violence, abortion, and sexual assault (Wolfe & Kimerling, 1997). For each positively endorsed item, the participant also responded to the question “How much has this event affected your life in the past year?” (i.e., during the entire pregnancy and immediate pre-conception period). We created a weighted score by summing the participant’s response (ranging from 1 = not at all to 5 = extremely) to this follow-up question about stress perception. The weighted score has a possible range from 0–150 and is designed to reflect both the frequency of stressful and traumatic events in the woman’s lifetime and her appraisal of the impact of these events in the prior year.

Maternal interpersonal trauma during childhood

Participants completed the Childhood Trauma Questionnaire (CTQ) during the same 24–27 week gestation study visit as the LSC-R. The CTQ is a retrospective, self-report index of the severity of traumatic childhood experiences through age 11 years across several types of maltreatment (Bernstein & Fink, 1998). We created three dichotomous domain scores (emotional abuse, physical abuse, sexual abuse) based on the participant’s response to seven questions about her own childhood experiences. Specifically, we scored emotional abuse as occurring if the participant endorsed either of the items: “Someone in my family yelled and screamed at me” or “People in my family said hurtful or insulting things to me” as sometimes, often, or very often true. We considered physical abuse as occurring if the participant endorsed any of the items: “People in my family hit me so hard that it left me with bruises and marks”, “The punishments I received seemed cruel”, or “I was punished with a belt, a board, a cord, or some other hard object” as sometimes, often, or very often true. Finally, we scored sexual abuse as occurring if the participant endorsed either of the items: “I believe that I was sexually abused by someone in my family” or “Someone molested me in the family” as sometimes, often, or very often true. The high validity and reliability of the CTQ has been demonstrated among multiple clinical and non-clinical samples (Bernstein & Fink, 1998; Scher, Stein, Asmundson, McCreary, & Forde, 2001).

Infant gestational age at birth

Information on gestational age at birth was extracted from newborn medical records. Gestational age was calculated based on: 1) date of delivery and self-reported last menstrual period and 2) ultrasound estimates from the first-trimester examination. If the discrepancy between the two sources was greater than two weeks, we used data from obstetrical estimates. We defined preterm birth as gestational age < 37 weeks.

Covariates

Information on sociodemographic and lifestyle characteristics was ascertained using questionnaires administered at enrollment and the third trimester. Participants self-reported Hispanic ethnicity and racial identity. For statistical analyses, we coded participants into one of three categories: 1) Black, Black Hispanic, African American, Haitian/Caribbean, 2) White non-Hispanic, or 3) White Hispanic, other Hispanic. We ascertained maternal age (continuous in years), parity (primiparous vs. multiparous), relationship status (single vs. married/living with partner), and education level (less than a high school degree vs. high school degree or more) by questionnaire. Maternal exposure to cigarette smoke during pregnancy was defined as self-reported active smoking (any) or exposure to environmental tobacco smoke for ≥ 1 hour per week during pregnancy. We calculated maternal pre-pregnancy body mass index (BMI, kg/m2) based on self-reported height and weight ascertained at enrollment.

We used Directed Acyclic Graph (DAG) theory to select covariates for inclusion in multivariable models based on substantive knowledge and published literature (Figure 2). We adjusted all models for maternal age, education, and race/ethnicity, which were sufficient to close biasing paths between STAXI-2 scores and PTB based on the conditional dependencies encoded by our DAG. In secondary analyses, we additionally adjusted for maternal cigarette exposure, parity, relationship status, and pre-pregnancy BMI to investigate the robustness of our selected model and consistency of findings across different covariate adjustment sets. We assumed the same DAG applied when examining main effects of LSC-R and CTQ scores and thus adjusted for the same covariate sets in these models.

Figure 2.

Figure 2.

Open in a new tab

Directed Acyclic Graph of assumed dependencies between maternal anger expression, preterm birth, and other socioeconomic, lifestyle, and health-related factors. Variables shaded in gray are those in the minimally sufficient adjustment set that would allow identifying any confounded effect of maternal anger expression on preterm birth. DAG constructed using DAGitty v2.3.

Statistical Analysis

We visually inspected descriptive characteristics of the study sample using histograms and boxplots. We used a modified Poisson regression model with robust error variance to estimate the relative risk (RR) of PTB associated with a 1-point increase in continuous STAXI-2 AX-I and AX-O scores, considered in separate models (Zou, 2004). We also examined associations between maternal lifetime stress, as indicated by continuous LSC-R scores, and childhood sexual, emotional, or physical abuse, as indicated by dichotomous CTQ scores, on the risk of PTB, in four separate modified Poisson regression models. Finally, we used cross-product terms to examine interactions between AX-I or AX-O and 1) LSC-R, 2) CTQ sexual abuse, 3) CTQ emotional abuse, and 4) CTQ physical abuse scores in separate models. All statistical analyses were performed using SAS version 9.4 (Cary, NC).

Results

Sample characteristics

Gestational length and scores on the STAXI-2, LSC-R, and CTQ were similar between participants included and excluded from the analysis. Participants excluded due to missing data were less likely to be exposed to cigarette smoke during pregnancy (17% vs. 29%), less likely to be White (10% vs. 21%), and more likely to be Hispanic (46% vs. 35%); we identified no other sociodemographic differences between the women included and the women excluded. Table 1 provides sample descriptive characteristics. The incidence of PTB was 11% and was similar across strata of maternal race/ethnicity (White: 8%, Black: 10%, Hispanic: 12%).

Table 1.

Sample characteristics. Values are median±interquartile range or n (%).

Maternal age 28.7±9.2
Race/ethnicity
 Black 281 (45)
 Hispanic 217 (35)
 White 130 (21)
<High school degree 144 (23)
Single relationship status 204 (32)
Primiparae 233 (37)
Any cigarette smoke exposurea 186 (30)
Pre-pregnancy body mass index 25.5±8.4
Gestational age (weeks) 39±2
Preterm birth (<37 weeks) 67 (11)
STAXI-2 Anger-out 15±5
STAXI-2 Anger-in 14±6
Life Stressor Checklist-Revised 9±12
CTQ emotional abuse 395 (63)
CTQ physical abuse 239 (38)
CTQ sexual abuse 59 (9)
Open in a new tab

Abbreviations: CTQ: Childhood Trauma Questionnaire; STAXI-2: State-Trait Anger Expression Inventory-2.

a

Active smoking or exposure to environmental tobacco smoke for ≥ 1 hour per week during pregnancy.

STAXI-2 AX-I (mean±SD: 14.3±4.2, range: 8–32) and AX-O (15.6±4.2, 8–32) scores were approximately normally distributed and moderately correlated (Spearman r=0.32). On average, AX-I and AX-O scores were higher among younger (AX-I: p=0.0008, AX-O p=0.009), single (AX-I: p<0.0001, AX-O: p<0.0001), minority (AX-I: p<0.0006, AX-O: p<0.0001) women who smoked or reported exposure to environmental cigarette smoke (AX-I: p=0.07, AX-O: p<0.0001) (Figure 3). AX-I scores were also higher among women with less than a high school education (p<0.0001). LSC-R scores were positively skewed (median±IQR: 9±12, range: 0–100) and positively correlated with both AX-I (Spearman r=0.23) and AX-O (Spearman r=0.37) scores. A majority of participants reported experiencing emotional abuse during childhood (n=395, 63%), approximately one-third reported physical abuse (n=239, 38%), and fewer than 10% reported sexual abuse (n=59, 9%). The correlations between continuous AX-I and AX-O scores with dichotomous childhood emotional (AX-O = 0.39, AX-I = 0.19), physical (AX-I = 0.19, AX-O = 0.20), and sexual (AX-I = 0.26, AX-O = 0.24) abuse were low to moderate (all rank biserial).

Figure 3.

Figure 3.

Open in a new tab

Mean±standard deviation STAXI-2 Anger-In Expression and Anger-Out Expression scores among women overall and stratified by sociodemographic and lifestyle characteristics.

Associations of trait anger expression, lifetime stress, and childhood trauma with PTB

In adjusted models, higher anger expression was associated with an increased risk of PTB for both the AX-O (adjusted risk ratio [aRR]=1.07, 95% confidence interval [CI]: 1.02, 1.12) and AX-I (aRR=1.06, 95% CI: 1.01, 1.11) scales (Table 2). We also detected an association between LSC-R scores and PTB (aRR= 1.02, 95% CI: 1.01, 1.03). In contrast, we did not detect significant relationships between childhood emotional, physical, or sexual abuse and PTB. We did not observe significant interactions between AX-I (p-interaction: AX-I ✕ LSC-R = 0.58; AX-I ✕ emotional abuse = 0.92; AX-I ✕ physical abuse = 0.76; AX-I ✕ sexual abuse = 0.15) or AX-O (p-interaction: AX-O ✕ LSC-R = 0.66; AX-O ✕ emotional abuse = 0.40; AX-O ✕ physical abuse = 0.21; AX-O ✕ sexual abuse = 0.13) scores and lifetime stress or childhood trauma.

Table 2.

Preterm birth in relation to maternal trait anger expression, lifetime stress, and childhood trauma, in separate models, among mother-newborn pairs enrolled in the PRISM pregnancy cohort (n=628).

Model 1 Model 2 Model 3
RR (95% CI) RR (95% CI) RR (95% CI)
STAXI-2 Anger-out 1.06 (1.02, 1.11) 1.07 (1.02, 1.12) 1.06 (1.01, 1.11)
STAXI-2 Anger-in 1.05 (1.01, 1.10) 1.06 (1.01, 1.11) 1.05 (1.00, 1.10)
LSC-R lifetime stress 1.02 (1.01, 1.03) 1.01 (1.01, 1.03) 1.02 (1.00, 1.03)
CTQ sexual abuse 0.95 (0.43, 2.10) 0.99 (0.44, 2.20) 0.89 (0.41, 1.96)
CTQ physical abuse 1.17 (0.74, 1.85) 1.13 (0.69, 1.85) 1.06 (0.63, 1.76)
CTQ emotional abuse 1.49 (0.99, 2.47) 1.54 (0.91, 2.60) 1.41 (0.81, 2.45)
Open in a new tab

Abbreviations: CTQ: Childhood Trauma Questionnaire; LSC-R: Life Stressor Checklist-Revised; STAXI-2: State-Trait Anger Expression Inventory-2

Model 1: unadjusted

Model 2: adjusted for maternal age, race/ethnicity, and education

Model 3: adjusted for Model 2 variables plus cigarette smoke exposure (active smoking or exposure to ≥ 1 hour per week during pregnancy), parity, relationship status, and pre-pregnancy body mass index

Discussion

In this study, greater anger expression was associated with younger maternal age, single relationship status, and racial/ethnic minority status and was positively correlated with lifetime traumatic and non-traumatic experiences. We found that higher inward and outward trait anger was associated with PTB even when adjusting for potential confounders, albeit effect sizes were small. We also found a positive association between lifetime stress and PTB, but did not detect associations with childhood abuse subscales. We did not find significant interactions between anger expression and lifetime stress or childhood abuse on the risk of PTB.

Anger is a common emotion with important functional properties, including signaling the occurrence of harmful or unjust events. However, frequent and intense experiences of anger can have pathophysiological consequences for health. Short-lasting anger (state anger) has been associated with transient increases in heart rate and blood pressure (Lupis, Lerman, & Wolf, 2014); likewise, chronic, longer-lasting tendencies to experience anger when provoked (trait anger) has been associated with the development of chronic disorders, including coronary heart disease and hypertension (Rutledge & Hogan, 2002; Suls, 2013). Experimental research also indicates the expression of anger, not only its experience, is associated with increased blood pressure reactivity, elevated resting blood pressure, and hypertension (Everson, Goldberg, Kaplan, Julkunen, & Salonen, 1998). In turn, chronic hypertension increases the risk for pregnancy-related complications, with a recent systematic review reporting a PTB incidence rate of 28% among women with chronic hypertension (Bramham et al., 2014).

Anger is a stress response mediated by the HPA axis. Feelings of anger and hostility have been associated with augmented HPA axis activity, as indicated by anger-elicited elevations in cortisol output (Steptoe, Cropley, Griffith, & Kirschbaum, 2000). In turn, elevated cortisol during pregnancy is a risk factor for PTB, putatively via placental upregulation of corticotrophin releasing hormone (CRH) (Wadhwa, Porto, Garite, Chicz-DeMet, & Sandman, 1998). CRH, known as the “placental clock,” increases throughout pregnancy and acts as a trigger for term parturition by stimulating the production and output of prostaglandins, which promote myometrial contractility (Olson, 2003). Given that CRH is significantly elevated among women experiencing preterm labor (Wadhwa et al., 1998), it is possible that altered CRH production in part mediates stress and/or anger-associated PTB; however, research investigating this hypothesis remains limited.

Immune activation could also contribute to associations between anger expression and PTB (Areia, Moura, Mota-Pinto, & CRD, 2019). Anger and the memory of anger-triggering events have been associated with upregulation of pro-inflammatory cytokines, including plasma interleukin-6, tumor necrosis factor-alpha, and c-reactive protein levels (Suarez, Boyle, Lewis, Hall, & Young, 2006). Pro-inflammatory cytokines orchestrate multiple aspects of the labor process, including stimulation of uterine contractility (Areia et al., 2019). We are not aware of research that has investigated anger-related immune activation in relation to PTB; however, other factors that enhance inflammation in pregnancy have been associated with PTB. For example, infection during pregnancy has been shown to activate toll-like receptors and stimulate production of inflammatory cytokines, resulting in preterm parturition (Cappelletti, Della Bella, Ferrazzi, Mavilio, & Divanovic, 2016). The putative role of the immune system in PTB is also supported by murine models showing that localized intrauterine inflammation results in 100% preterm delivery via toll-like receptor-4 activity (Elovitz, Wang, Chien, Rychlik, & Phillippe, 2003).

Our finding of a small, but positive association between maternal lifetime stress and PTB is consistent with previous research that supports a relation between maternal psychosocial stress before and/or during pregnancy with PTB, even after accounting for other behavioral, socio-demographic, and medical risks. For example, a quantitative synthesis of reviews on this topic suggests that women reporting higher psychosocial stress have approximately a 25–60% increased risk for PTB compared to women with lower levels of stress (Wadhwa, Entringer, Buss, & Lu, 2011). While a comprehensive discussion of all available findings on maternal stress and PTB is beyond the scope of this discussion, we refer readers to two reviews that summarize accumulating evidence on this topic (Entringer, Buss, & Wadhwa, 2010; Wadhwa et al., 2011).

In contrast to findings with maternal lifetime stress and trait anger expression, we did not observe associations between maternal trauma during childhood and the risk of PTB. One explanation may be the relatively long latency period between childhood and pregnancy, which may make it difficult to detect true causal agents due to competing exposures and intervening experiences over time. Additionally, we may have been underpowered to detect associations due to the low number of participants with both trauma and PTB (sexual abuse: n=6, physical abuse: n=28, emotional abuse: n=48). Further, our dichotomous measures did not capture severity and/or chronicity of the abuse experiences, which may be important considerations.

Notable study strengths include the prospective design, the use of well-validated measures for childhood trauma, lifetime stress, and anger, and the ability to adjust for a number of potential confounders. The PRISM cohort is comprised largely of lower-income, minority women living in a major U.S. urban metropolitan area. This is a demographic that is at risk for heightened exposure to stress/trauma and reduced access to buffering resources (APA, 2017). We therefore expect that scores on measures of stress and trauma in this sample may be higher than the general U.S. population, which could limit the generalizability of these findings. However, we note that the prevalence rates of physical (38%) and sexual abuse (9%) in PRISM are within the range reported by nonclinical samples in the United States, which vary depending on definition and show a prevalence rate in women of 6–40% for physical abuse and 7–36% for sexual abuse (Springer, Sheridan, Kuo, & Carnes, 2003). The prevalence rate (63%) of emotional abuse in PRISM may be higher than that reported by some studies as we included positively endorsed ‘sometimes’ responses in our definition based on literature linking even occasional experiences of abuse during childhood with physical and psychological health outcomes (Anda et al., 2006; Jonson-Reid, Kohl, & Drake, 2012). Additionally, the incidence of PTB was similar in the present study to that reported nationally in the United States (Pruessner, Hellhammer, & Kirschbaum, 1999). Several limitations also warrant comment. We were unable to separate medically-indicated from spontaneous PTB. These phenotypes are characterized by different underlying etiologies; thus, how they relate to maternal anger and stress may vary. If anger expression is only truly linked to spontaneous or medically-indicated PTB, then our inability to distinguish between potentially distinct physiological pathways may have biased our results towards the null. We were unable to test for mediation by hypertensive disorders given the low incidence of chronic hypertension (5.0%), gestational hypertension (6.3%), pre-eclampsia (7.2%), and eclampsia (0.2%) in our generally healthy, community-based sample. As with any study that relies on medical record data for gestational age estimation, misclassification of preterm birth, which we expect would be non-differential, is also possible. We attempted to minimize misclassification by relying on best obstetrical estimates when self-reported last menstrual period and ultrasound obstetrical estimates differed by two weeks.

Preterm birth represent one of the most significant problems in maternal and child health (Wadhwa et al., 2011). The final months of pregnancy are a critical developmental period during which growth accelerates and organ systems mature. Birth before this process is complete is associated with an increased risk of mortality and serious neonatal complications that carry lifelong consequences for health and well-being. Identifying modifiable risk factors for preterm delivery, including traumatic stress and emotional responding, can inform clinical practice agendas for addressing the problem of preterm birth and offers a strategy to improve maternal and child health. The American College of Obstetricians and Gynecologists recommends psychosocial screening during pregnancy (American College of & Gynecologists Committee on Health Care for Undeserved, 2006) and several studies suggest that anger management interventions that teach adaptive anger control and coping skills are effective for enhancing emotional and physical wellbeing (Blake & Hamrin, 2007). These programs aim to promote self-awareness about the types and function of anger and somatic experiences associated with anger and teach productive ways to express and cope with angry feelings (Candelaria, Fedewa, & Ahn, 2012). The specific techniques employed in anger management interventions, such as emotional awareness, relaxation training, cognitive restructuring, social skills development, and conflict resolution, are not contraindicated during pregnancy and may impact other health behaviors with potential benefits for pregnancy outcomes (e.g., reducing smoking and substance abuse). Moreover, improving maternal anger management and skills for coping with negative emotions could also translate to improved development for the mother’s own child, thereby potentially improving health and wellbeing across generations.

Expression of anger among pregnant women in this sample was associated with an increased risk of PTB, although the effect size was small. Future research is needed to corroborate these findings and explore putative underlying mechanisms, including hypertension-related pathways, altered maternal HPA axis activity, and stimulation of a pro-inflammatory state during pregnancy. Ultimately, understanding how anger and other emotional and psychosocial factors amenable to intervention are linked to PTB can help inform the development of strategies to improve birth outcomes.

Acknowledgments

Funding: This work was supported by the National Institutes of Health [grant numbers: R01 HL095606, R01 HL114396, and P30 ES023515]. During preparation of this manuscript, WC was supported by T32 HD049311.

References

  1. Adam EK (2006). Transactions among adolescent trait and state emotion and diurnal and momentary cortisol activity in naturalistic settings. Psychoneuroendocrinology, 31(5), 664–679. 10.1016/j.psyneuen.2006.01.010 [DOI] [PubMed] [Google Scholar]
  2. American College of, O., & Gynecologists Committee on Health Care for Undeserved, W. (2006). ACOG Committee Opinion No. 343: psychosocial risk factors: perinatal screening and intervention. Obstet Gynecol, 108(2), 469–477. 10.1097/00006250-200608000-00046 [DOI] [PubMed] [Google Scholar]
  3. Anda RF, Felitti VJ, Bremner JD, Walker JD, Whitfield C, Perry BD, … Giles WH (2006). The enduring effects of abuse and related adverse experiences in childhood. A convergence of evidence from neurobiology and epidemiology. Eur Arch Psychiatry Clin Neurosci, 256(3), 174–186. 10.1007/s00406-005-0624-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. APA. (2017). Stress and health disparities: Contexts, mechanisms, and interventions among racial/ethnic minority and low-socioeconomic status populations. Retrieved from http://www.apa.org/pi/health-disparities/resources/stress-report.aspx
  5. Areia AL, Moura P, Mota-Pinto A, & CRD, P. N. m. (2019). The role of innate immunity in spontaneous preterm labor: A systematic review. J Reprod Immunol, 136, 102616. 10.1016/j.jri.2019.102616 [DOI] [PubMed] [Google Scholar]
  6. Bernstein D, & Fink L (1998). Childhood Trauma Questionnaire: A retrospective self-report. San Antonio, TX: The Psychological Corporation. [Google Scholar]
  7. Blake CS, & Hamrin V (2007). Current approaches to the assessment and management of anger and aggression in youth: a review. J Child Adolesc Psychiatr Nurs, 20(4), 209–221. 10.1111/j.1744-6171.2007.00102.x [DOI] [PubMed] [Google Scholar]
  8. Bramham K, Parnell B, Nelson-Piercy C, Seed PT, Poston L, & Chappell LC (2014). Chronic hypertension and pregnancy outcomes: systematic review and meta-analysis. BMJ, 348, g2301. 10.1136/bmj.g2301 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Candelaria A, Fedewa A, & Ahn S (2012). The effects of anger management on children’s social and emotional outcomes: a meta-analysis. School Psychology International, 33(6), 596–614. [Google Scholar]
  10. Cappelletti M, Della Bella S, Ferrazzi E, Mavilio D, & Divanovic S (2016). Inflammation and preterm birth. J Leukoc Biol, 99(1), 67–78. 10.1189/jlb.3MR0615-272RR [DOI] [PubMed] [Google Scholar]
  11. Christiaens I, Hegadoren K, & Olson DM (2015). Adverse childhood experiences are associated with spontaneous preterm birth: a case-control study. BMC Med, 13, 124. 10.1186/s12916-015-0353-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Deschenes SS, Dugas MJ, Fracalanza K, & Koerner N (2012). The role of anger in generalized anxiety disorder. Cogn Behav Ther, 41(3), 261–271. 10.1080/16506073.2012.666564 [DOI] [PubMed] [Google Scholar]
  13. Elovitz MA, Wang Z, Chien EK, Rychlik DF, & Phillippe M (2003). A new model for inflammation-induced preterm birth: the role of platelet-activating factor and Toll-like receptor-4. Am J Pathol, 163(5), 2103–2111. 10.1016/S0002-9440(10)63567-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Entringer S, Buss C, & Wadhwa PD (2010). Prenatal stress and developmental programming of human health and disease risk: concepts and integration of empirical findings. Curr Opin Endocrinol Diabetes Obes, 17(6), 507–516. 10.1097/MED.0b013e3283405921 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Everson SA, Goldberg DE, Kaplan GA, Julkunen J, & Salonen JT (1998). Anger expression and incident hypertension. Psychosom Med, 60(6), 730–735. 10.1097/00006842-199811000-00014 [DOI] [PubMed] [Google Scholar]
  16. Field T, Diego M, Hernandez-Reif M, Salman F, Schanberg S, Kuhn C, … Bendell D (2002). Prenatal anger effects on the fetus and neonate. J Obstet Gynaecol, 22(3), 260–266. 10.1080/01443610220130526 [DOI] [PubMed] [Google Scholar]
  17. Jonson-Reid M, Kohl PL, & Drake B (2012). Child and adult outcomes of chronic child maltreatment. Pediatrics, 129(5), 839–845. 10.1542/peds.2011-2529 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Karlamangla AS, Merkin SS, Almeida DM, Friedman EM, Mogle JA, & Seeman TE (2019). Early-Life Adversity and Dysregulation of Adult Diurnal Cortisol Rhythm. J Gerontol B Psychol Sci Soc Sci, 74(1), 160–169. 10.1093/geronb/gby097 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Laszlo KD, Janszky I, & Ahnve S (2010). Anger expression and prognosis after a coronary event in women. Int J Cardiol, 140(1), 60–65. 10.1016/j.ijcard.2008.10.028 [DOI] [PubMed] [Google Scholar]
  20. Lupis SB, Lerman M, & Wolf JM (2014). Anger responses to psychosocial stress predict heart rate and cortisol stress responses in men but not women. Psychoneuroendocrinology, 49, 84–95. 10.1016/j.psyneuen.2014.07.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Martin J, Hamilton B, Osterman J, Driscoll A, & Drake P (2018). Births: Final Data for 2016. National vital statistics reports: from the Centers for Disease Control and Prevention, National Center for Health Statistics. Nat Vital Stat Syst, 67(1). [PubMed] [Google Scholar]
  22. Moster D, Lie RT, & Markestad T (2008). Long-term medical and social consequences of preterm birth. N Engl J Med, 359(3), 262–273. 10.1056/NEJMoa0706475 [DOI] [PubMed] [Google Scholar]
  23. Muglia LJ, & Katz M (2010). The enigma of spontaneous preterm birth. N Engl J Med, 362(6), 529–535. 10.1056/NEJMra0904308 [DOI] [PubMed] [Google Scholar]
  24. OECD. (2018). Infant mortality rates (indicator).
  25. Olson DM (2003). The role of prostaglandins in the initiation of parturition. Best Pract Res Clin Obstet Gynaecol, 17(5), 717–730. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/12972010 [DOI] [PubMed] [Google Scholar]
  26. Polk DE, Cohen S, Doyle WJ, Skoner DP, & Kirschbaum C (2005). State and trait affect as predictors of salivary cortisol in healthy adults. Psychoneuroendocrinology, 30(3), 261–272. 10.1016/j.psyneuen.2004.08.004 [DOI] [PubMed] [Google Scholar]
  27. Pruessner JC, Hellhammer DH, & Kirschbaum C (1999). Burnout, perceived stress, and cortisol responses to awakening. Psychosom Med, 61(2), 197–204. 10.1097/00006842-199903000-00012 [DOI] [PubMed] [Google Scholar]
  28. Rutledge T, & Hogan BE (2002). A quantitative review of prospective evidence linking psychological factors with hypertension development. Psychosom Med, 64(5), 758–766. 10.1097/01.psy.0000031578.42041.1c [DOI] [PubMed] [Google Scholar]
  29. Scher CD, Stein MB, Asmundson GJ, McCreary DR, & Forde DR (2001). The childhood trauma questionnaire in a community sample: psychometric properties and normative data. J Trauma Stress, 14(4), 843–857. 10.1023/A:1013058625719 [DOI] [PubMed] [Google Scholar]
  30. Spielberger D (1999). STAXI-2 state trait anger expression inventory-2, professional manual. Florida: PAR. [Google Scholar]
  31. Springer KW, Sheridan J, Kuo D, & Carnes M (2003). The long-term health outcomes of childhood abuse. An overview and a call to action. J Gen Intern Med, 18(10), 864–870. 10.1046/j.1525-1497.2003.20918.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Staneva A, Bogossian F, Pritchard M, & Wittkowski A (2015). The effects of maternal depression, anxiety, and perceived stress during pregnancy on preterm birth: A systematic review. Women Birth, 28(3), 179–193. 10.1016/j.wombi.2015.02.003 [DOI] [PubMed] [Google Scholar]
  33. Steptoe A, Cropley M, Griffith J, & Kirschbaum C (2000). Job strain and anger expression predict early morning elevations in salivary cortisol. Psychosom Med, 62(2), 286–292. 10.1097/00006842-200003000-00022 [DOI] [PubMed] [Google Scholar]
  34. Suarez EC, Boyle SH, Lewis JG, Hall RP, & Young KH (2006). Increases in stimulated secretion of proinflammatory cytokines by blood monocytes following arousal of negative affect: the role of insulin resistance as moderator. Brain Behav Immun, 20(4), 331–338. 10.1016/j.bbi.2005.09.005 [DOI] [PubMed] [Google Scholar]
  35. Suls J (2013). Anger and the heart: perspectives on cardiac risk, mechanisms and interventions. Prog Cardiovasc Dis, 55(6), 538–547. 10.1016/j.pcad.2013.03.002 [DOI] [PubMed] [Google Scholar]
  36. Wadhwa PD, Entringer S, Buss C, & Lu MC (2011). The contribution of maternal stress to preterm birth: issues and considerations. Clin Perinatol, 38(3), 351–384. 10.1016/j.clp.2011.06.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Wadhwa PD, Porto M, Garite TJ, Chicz-DeMet A, & Sandman CA (1998). Maternal corticotropin-releasing hormone levels in the early third trimester predict length of gestation in human pregnancy. Am J Obstet Gynecol, 179(4), 1079–1085. 10.1016/s0002-9378(98)70219-4 [DOI] [PubMed] [Google Scholar]
  38. Wei S, Xiao X, Wang J, Sun S, Li Z, Xu K, … Qiao M (2017). Impact of anger emotional stress before pregnancy on adult male offspring. Oncotarget, 8(58), 98837–98852. 10.18632/oncotarget.22007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Wolfe J, & Kimerling R (1997). Gender issues in the assessment of posttraumatic stress disorder. In Wilson & Keane TM (Eds.), Assessing psychological trauma and PTSD (pp. 192–238). New York, NY, US: Guilford Press. [Google Scholar]
  40. Wolfe J, Kimerling R, Brown PJ, Chrestman KR, & Levin K (1996). Psychometric review of the life stressor checklist-revised. Lutherville, MD: Sidran Press. [Google Scholar]
  41. Zou G (2004). A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol, 159(7), 702–706. 10.1093/aje/kwh090 [DOI] [PubMed] [Google Scholar]

RESOURCES