Intimate Partner Violence and Psychological Distress Among Emerging Adult Women: A Bidirectional Relationship

Abstract

Background: Intimate partner violence (IPV) and psychological distress (PD) are major public health concerns among emerging adult women. Emerging adulthood presents a complex set of new experiences and challenges that pose a risk to normative development. In particular, an increased prevalence of IPV and PD during this time period may lead to long-term health consequences.

Methods: Data from the Relationship Dynamics and Social Life study, a longitudinal study of a racially and socioeconomically diverse population-representative random sample of 726 partnered women, aged 18–19, residing in a Michigan county, and followed for 2.5 years, were used to investigate the relationship between IPV and PD. Logistic regression models predicted each measure of PD (depression, stress, loneliness, self-esteem) as a function of past IPV (none, psychological violence only, any physical violence), and multinomial logistic regression models predicted subsequent weekly IPV as a function of each measure of PD.

Results: PD and IPV were prevalent among emerging adult women. Past psychological IPV was associated with experiencing all four distress measures. Past physical IPV was also associated with depression, stress, and loneliness, but not self-esteem. Women with each PD were more likely to subsequently experience psychological violence, and women who reported stress were more likely to subsequently experience any physical violence.

Conclusions: The IPV–PD relationship is bidirectional. Women who experienced past IPV were more likely to report PD. Conversely, women who experienced PD were at a greater risk of subsequent IPV.

Background and Significance

Intimate partner violence (IPV) and psychological distress (PD) are major public health issues among emerging adult women. IPV is any physical, sexual, or psychological form of violence against an intimate partner. PD is a general term used to describe symptoms of mental health outcomes such as depression and anxiety. Arnett proposed that ages 18–25 constitute an unique developmental stage he termed emerging adulthood,¹ where persons in their late teens and early twenties view themselves as having left adolescence but not yet entered into young adulthood. Characterized by volatility and uncertainty, emerging adulthood presents a complex set of new experiences and challenges that pose a risk to normative development. In particular, the increased prevalence of IPV and PD during emerging adulthood can lead to severe health consequences.

Intimate partner violence

According to the CDC's latest National Intimate Partner and Sexual Violence Survey, almost one-third of women have experienced some form of violence perpetrated by an intimate partner; of these women, ∼71% first experienced IPV between the ages of 18 and 24.² IPV has many negative health consequences, including chronic pain, migraines, hypertension, gastrointestinal disorders, gynecological issues, and/or sexually transmitted infections.³ Between 2010 and 2012, nearly 10% of U.S. women had been raped by an intimate partner in their lifetime, 16.9% experienced some other form of sexual violence, 24.3% reported severe physical violence, 10.7% reported stalking, and 48.4% reported psychological violence.²

Although psychological violence affects almost twice as many women than physical violence, it is often overlooked or minimized because it does not leave visible injury.⁴ Some researchers posit that women may have greater difficulty recovering from psychological violence as opposed to physical violence due to its detrimental effect on their overall self-identity.⁵ Psychological violence can be perpetrated through various forms, including threats, verbal abuse, accusations, and/or manipulation.^4–6 Women who suffered from psychological IPV reported subsequent loss of individuality, feelings of helplessness, and PD.⁶

Psychological distress

The foremost psychological consequences of IPV include depression, post-traumatic stress disorder, and substance use disorders.³ The current study concentrates on four measures of PD: depression, stress, loneliness, and self-esteem. Depression during emerging adulthood has been linked to adverse outcomes, including diminished self-esteem, lower educational attainment, and impaired work and social functioning.⁷ PD has also been associated with negative health behaviors, such as increased smoking and alcohol consumption.⁸ There is a two- to threefold increased risk of major depression in women who experience IPV.⁹ Moreover, women who are depressed often have decreased social support networks and are less likely to leave violent relationships,⁷ which may also be attributed to learned helplessness.¹⁰

Stress is associated with increased depressive symptoms,¹¹ increased anxiety, and greater interpersonal difficulties. Among a sample of emerging adult women, nearly twice as many experienced a depressive episode due to interpersonal stress with romantic partners, peers, or family than non-interpersonal stress such as academic or health problems.¹¹ The challenges of these daily stressors combined with new responsibilities strain mental well-being, and exposure to violence further disrupts stress adaptation. Loneliness, particularly when combined with a lack of perceived social support, has detrimental consequences on one's ability to handle adversity.¹² Overall, the isolative nature of IPV diminishes women's self-esteem and induces feelings of worthlessness that can exacerbate depression.⁵

Individual and relationship factors

Sociodemographic characteristics and adolescent experiences related to sex and pregnancy are associated with an increased risk of experiencing IPV and PD. African American women experience more severe and frequent IPV compared with White women.¹³ Socioeconomic status is one of the strongest risk factors for IPV victimization.¹³ Low-income women are less likely to have access to resources and more likely to participate in risky behaviors such as alcohol and drug abuse; thus, financial dependence on violent partners may pressure women to stay in violent relationships.¹⁴ Low educational attainment is also correlated with greater IPV frequency.¹³

In regard to adolescent experiences related to sex and pregnancy, multiple sexual partners in adolescence has been linked to increased IPV risk in emerging adulthood.¹⁵ Women who engaged in casual sex were more likely to experience PD, often due to sexual regret.¹⁶ IPV increases the incidence of unprotected sex as intimate partners may utilize violence to interfere with contraception, resulting in unintended pregnancy.¹⁷ Moreover, IPV during pregnancy catalyzes detrimental outcomes, including preterm birth and low birth weight.¹⁸

Relationship seriousness is also associated with IPV. Increased cohabitation in emerging adult intimate relationships has been linked to higher incidences of violence.¹⁹ One theory is that cohabiting couples have longer relationships and share a residence, which may include greater opportunities for violent experiences to arise over time.¹⁹ Cohabitors have less defined roles than married couples and may be less able to develop appropriate conflict resolution strategies.¹⁹ While dating couples are also less committed to their relationships, they experience less violence because cohabitation increases the exposure to violence compared with dating couples who do not live together.¹⁹

Developing and maintaining healthy intimate relationships is an important task of emerging adulthood.¹ However, >4 out of 10 emerging adults have gone through a break-up and have gotten back together in their current or most recent relationship.²⁰ This relationship instability can incite loneliness and depression. In addition, previous experience with IPV may normalize or desensitize women to acts of abuse. In a qualitative study by Valdez, Lim, and Lilly, women in violent relationships rationalized abuse through desires for love or companionship.²¹ The debilitating effects of violence on a woman's psychological well-being increase the expectation for violence in future relationships,²¹ illuminating the potentially reciprocal relationship between IPV and PD.

There remains a paucity of literature evaluating the relationship between IPV, PD, and its impact on the distinct development of emerging adult women. Data limitations hinder the ability to account for the potentially reciprocal relationship between IPV and PD, as IPV may increase the risk of becoming distressed and, alternatively, distress may increase the risk of experiencing violence. Given this potential reciprocal IPV–PD relationship, we first examined the influence of past IPV on PD, which consisted of four measures: depression, stress, loneliness, and self-esteem. Second, we investigated the influence of PD on the incidence of subsequent IPV. For the purposes of this study, we focused on psychological and physical IPV.

Based on the literature, we hypothesized that past IPV leads to an increased risk of experiencing PD, and that PD increases the risk of subsequent IPV. Our findings will be beneficial to the early identification and intervention of IPV, which is crucial for public health.

Methods

Data

The Relationship Dynamic and Social Life (RDSL) study began with a population-representative sample of 1,003 respondents randomly selected from a list of 18- and 19-year-old women in a single Michigan county. Women were selected from the state driver's license and personal identification card databases. (Out of the original sample of 1,194 individuals, 1,085 could be located (93%), but after removing the 82 potential respondents who declined, the final sample size was 1,003 respondents (84% of the original sample). In all, 109 individuals were nonresponders.) Professional interviewers conducted baseline face-to-face interviews between March 2008 and July 2009 to assess sociodemographic characteristics, relationship characteristics, and mental health, which we term PD.

At the conclusion of the baseline interview, respondents were invited to participate in the weekly survey component that covered a 2.5-year follow-up period. The weekly surveys collected measures of relationship characteristics such as commitment and violence via 5-minute phone or web surveys. Respondents received $1 per weekly survey with $5 bonuses for on-time completion of five consecutive weekly surveys. Of the 1,003 women who completed the baseline interview, 95% (N = 953) participated in weekly surveys. The follow-up component concluded in January 2012.

The analytic sample for this investigation comprises the 726 respondents who identified a relationship at the baseline interview and all the weekly surveys in which these respondents were in relationships (N = 23,600). (Overall, the study included 58,594 weekly interviews from 953 young women. This represents 47% of the total surveys that would have been completed if all respondents completed a survey every 7 days for 2.5 years. All surveys completed 2 or more weeks after the prior survey were adjusted to refer to the prior week; thus, there is very little missing data. We refer to the surveys as “weekly” because that is approximately the median [8 days] and the mode [7 days] of the number of days between surveys. Ninety-one percent of all surveys were completed within 14 days.)

Measures

Intimate partner violence

We conceptualized IPV as a categorical measure of no violence, psychological violence only, or any physical violence. Respondents were asked if they had fought or had any arguments with their partner. Respondents who answered yes were then asked about specific types of fighting. Psychological violence was established by asking the respondent if her partner had sworn at her, called her names, insulted her, treated her disrespectfully, or threatened her with violence. Any physical violence was determined by asking the respondent if her partner had pushed, hit, or thrown something at her that could hurt. Physical violence may or may not include psychological violence.

At the baseline interview, respondents were asked if their partner had ever perpetrated only psychological or any physical violence. We refer to IPV before the baseline interview as past IPV, that is, “past” relative to PD. In the weekly journal surveys, respondents were asked if their partner had perpetrated only psychological or any physical violence since the last survey. We refer to IPV reported in each weekly journal after the baseline interview as subsequent IPV, that is, “subsequent” relative to PD.

For both measures, IPV is a mutually exclusive categorical variable that equals 1 if the respondent experienced psychological violence only, or 2 if the respondent experienced any physical violence by her partner. The reference category for violence is no violence, which is denoted as 0. We combined physical only and physical and psychological violence due to the small cell size for physical only and since our preliminary analyses found the two categories to be similar.

Psychological distress

Mental health questions were asked at the baseline interview only. We therefore refer to PD as baseline PD. We investigated four measures of PD: depression, stress, loneliness, and self-esteem. We coded depression and stress similarly to Hall et al.²²

Depression

The Center for Epidemiologic Studies–Depression Scale-5 (CES-D-5) is an abbreviated depression scale that assesses depressive symptoms over the previous week.²³ Depression was identified by the following symptoms: (a) felt like you could not shake off the blues, (b) felt depressed, (c) felt sad, (d) felt that life was not worth living, and (e) felt happy. The positively worded item (e) was reverse-coded. Responses were scored on a four-point scale: 0 = rarely or none of the time, 1 = some or little of the time, 2 = occasionally or moderate amount of time, and 3 = most or all of the time. Items were then summed for a total depression score. Scores ranged from 0 to 15; a higher score indicated a higher degree of symptoms. We used a standard score cut-off of ≥4 points on the CES-D-5 to denote moderate to severe depression.

Stress

The Perceived Stress Scale-4 (PSS-4) is an abbreviated scale that assesses the degree to which one appraises his or her life situations as stressful, unpredictable, uncontrollable, and overloading over the previous month.²⁴ Stress was established by assessing the respondents' perception of: (a) inability to control important things in life, (b) confidence in ability to solve problems, (c) felt that things were going your way, and (d) felt that difficulties cannot be overcome. Positively worded items (b) and (c) were reverse-coded. Responses were scored on a four-point scale: 0 = never, 1 = almost never, 2 = sometimes, 3 = fairly often, 4 = very often. Items were summed for a total score, which could range from 0 to 16, with higher scores indicating greater stress. We used a standard score cut-off of ≥9 points to denote moderate to severe stress.

Loneliness

The Revised UCLA Loneliness Scale assesses subjective feelings of loneliness and social isolation.²⁵ Respondents were asked how often (0 = never, 1 = rarely, 2 = sometimes, and 3 = often) they felt: (a) lack of companionship, (b) close to people, (c) left out, or (d) there are people you can turn to. Positively worded items (b) and (d) were reverse-coded. Items were summed for a total score, which could range from 0 to 12, with higher scores indicating greater symptoms of loneliness. We used a standard score cut-off of ≥6 points to denote moderate to severe loneliness.

Self-esteem

The Rosenberg Self-Esteem Scale is a global self-worth screen that measures both positive and negative feelings toward the self.²⁶ Self-esteem items were described as: (a) satisfied with self, (b) not much to be proud of, (c) feel like failure, and (d) positive toward self. Negatively worded items (a) and (d) were reverse-coded. These questions were then scored on a four-point Likert scale: 1 = strongly disagree, 2 = disagree, 3 = agree, and 4 = strongly agree. Items were summed for a total score, which could range from 1 to 16, with higher scores indicating greater self-esteem. We used a standard score cut-off of ≥13 points to denote high self-esteem.

Individual characteristics

We controlled for several self-reported sociodemographic characteristics measured at baseline, including age, race, childhood disadvantage, religious importance, education, and public assistance. Although only 18- and 19-year-old women were chosen for sampling, a small number turned 20 before they were located for the baseline interview. Race was dichotomized into African American versus not African American because the RDSL sample was predominantly White or African American.

An indicator of childhood disadvantage was created by summing the following four dichotomous measures: grew up with one biological parent only (no step-parent) or with another arrangement such as grandparents, an aunt, or other family member; biological mother was a teenager when she had her first child; mother's highest level of education was less than high school; and family received public assistance during childhood. Religious importance was categorized as not at all or somewhat important versus very important or more important than anything else.

Since the majority of RDSL respondents had not yet completed their education at the baseline interview, education was based on high school grade point average (GPA), which reflects recent school performance and should indicate future educational attainment. Public assistance was determined by assistance from WIC (Women, Infants and Children Program), FIP (Family Independence Program), cash welfare, or food stamps at age 18 or 19. We also controlled for four baseline measures of past adolescent experiences related to sex and pregnancy: age at first sexual intercourse, number of sexual partners, sexual intercourse without the use of birth control, and past pregnancies.

Relationship characteristics

Finally, we also controlled for relationship type.^† Relationship type was measured using several questions about marriage and engagement, cohabitation, commitment, and time spent together. Respondents reported whether they “spent a lot of time” with their partner (time-intensive). Commitment was gauged by whether the respondent and her partner “agreed to only have a special romantic relationship with each other, and no one else.” Respondents also reported whether they cohabited with their partner (whether they lived in a place that was “separate from where your partner lives”) and whether they were engaged to be married or married to their partner. (Questions about commitment and time spent together were not asked of respondents in married/engaged or cohabiting relationships.)

We then categorized relationship type using the following mutually exclusive hierarchical categories, which vary from week to week: (a) casual—uncommitted and less time-intensive, (b) nonexclusive dating—uncommitted but more time-intensive, (c) long distance—committed but less time-intensive (e.g., partner deployed in the military, lives far away, etc.), (d) exclusive dating—committed and more time-intensive, (e) cohabiting, and (f) married or engaged. The reference category for current relationship type is the least serious type, that is, casual.

Analytic strategy

We conducted all analyses using Stata/SE 14.0. First, we estimated logistic regressions predicting each measure of baseline PD as a function of past IPV. Second, we estimated multinomial logistic regressions to predict subsequent (journal) IPV as a function of each measure of baseline PD. Because the second regression used the weekly journal survey data, we utilized the cluster option to adjust the standard errors and account for observation clustering (relationship weeks) within respondents. In both sets of regressions, we coded IPV as no violence, psychological violence only, and any physical violence. In the latter set of regressions, we also controlled for past IPV. All models controlled for respondent characteristics, which included sociodemographics, adolescent experiences related to sex and pregnancy, and relationship type.

Results

Table 1 presents the distribution of women's baseline PD and past IPV experiences in relationships identified at baseline (N = 726 women), and follows these women throughout the study (N = 23,600 weeks). At baseline, 19% of women experienced IPV; 16% experienced psychological violence only, while 3% reported physical violence, which may or may not also include psychological violence. Twenty-six percent of women experienced depression, 24% reported stress, 27% reported loneliness, and 49% had low self-esteem at baseline.

Table 1.

Descriptive Statistics of Measures Used in the Analyses (N = 726 Women and N = 23,600 Weeks)

	Percentage/Mean (SD)
Baseline psychological distress (N = 726)
Depression	25.90
Stress	24.10
Loneliness	26.58
Low self-esteem	48.48
Past intimate partner violence (N = 726)
None	80.71
Only psychological	16.12
Any physical	3.17
Subsequent intimate partner violence (journal)
N = 726 (% women ever experienced)
Only psychological	33.06
Any physical	16.80
Sociodemographic characteristics (N = 726)
Age
18	40.36
19	51.38
20	8.26
African American	34.16
Highly religious	56.89
Childhood disadvantage	1.35 (1.13)
High school GPA	3.09 (0.62)
Receiving public assistance	30.30
Adolescent experiences related to sex and pregnancy (N = 726)
Age at first sex ≤16 years	60.33
Two or more sexual partners	67.90
Ever had sex without the use of birth control	56.61
One or more pregnancies	31.40
Baseline relationship type (N = 726)
Casual	4.13
Nonexclusive dating	6.75
Long-distance	9.78
Exclusive dating	52.48
Cohabiting	14.33
Married/engaged	12.53
Subsequent relationship type (journal) (N = 23,600)
Casual	4.64
Nonexclusive dating	3.33
Long-distance	17.32
Exclusive dating	31.38
Cohabiting	18.51
Married/engaged	24.82

GPA, grade point average.

Among the analytic sample, 34% identified as African American and 57% were highly religious. The childhood disadvantage mean was ∼1 on the 0–4 range. The average high school GPA was 3.09. Thirty percent of women were receiving public assistance at age 18 or 19. Sixty percent of women were ≤16 years when they first had sex, 68% reported two or more sexual partners, 57% had sex without the use of birth control, and 31% reported one or more pregnancies.

At baseline (N = 726), 4% of women were in casual relationships, 7% were nonexclusive dating, 10% were in long-distance relationships, 53% were exclusively dating, 14% were cohabiting, and 13% were married/engaged. Throughout the journals (N = 23,600), the women's relationship types shifted toward increased seriousness, reflected by the larger percentage of weeks in more serious relationships (i.e., exclusively dating and married/engaged).

Table 2 shows the logistic regression results of the effect of past IPV on baseline PD controlling for sociodemographic characteristics, adolescent experiences related to sex and pregnancy, and relationship type. Women who reported a history of psychological violence only were significantly more likely to experience depression, stress, and loneliness and were less likely to have high self-esteem compared with women without a history of violence. Women who reported a history of any physical violence were significantly more likely to experience depression, stress, and loneliness compared with women without a history of violence, with no significant differences in self-esteem net of controls.

Table 2.

Logistic Regression Results (Odds Ratios) of the Effect of Past Violence on Baseline Psychological Distress (N = 726 Women)

	Depression	Stress	Loneliness	High self-esteem
	ORs (CIs)
Past violence (ref: none)
Only psychological	1.66 (1.01–2.74)	2.29 (1.40–3.75)	1.60 (0.96–2.66)	0.55 (0.34–0.87)
Any physical	3.33 (1.33–8.36)	3.12 (1.26–7.75)	2.56 (1.02–6.40)	0.52 (0.21–1.30)
Sociodemographic characteristics
Age (ref: 18)
19	1.39 (0.96–2.03)	1.25 (0.86–1.83)	1.68 (1.15–2.46)	0.75 (0.54–1.04)
20	0.67 (0.30–1.50)	0.66 (0.30–1.48)	0.83 (0.38–1.82)	0.99 (0.55–1.78)
African American (ref: non-African American)	1.41 (0.92–2.17)	1.32 (0.85–2.05)	1.89 (1.24–2.90)	1.37 (0.94–2.02)
Highly religious	1.32 (0.88–1.96)	1.05 (0.71–1.57)	1.12 (0.75–1.66)	1.28 (0.91–1.78)
Childhood disadvantage	1.30 (1.09–1.56)	1.05 (0.87–1.26)	1.15 (0.96–1.37)	0.87 (0.75–1.03)
High school GPA	0.58 (0.43–0.78)	0.65 (0.48–0.88)	0.69 (0.51–0.93)	1.45 (1.11–1.89)
Receiving public assistance	0.75 (0.47–1.20)	0.64 (0.39–1.04)	1.08 (0.68–1.72)	0.74 (0.49–1.12)
Adolescent experiences related to sex and pregnancy
Age at first sex ≤16 years	1.67 (1.04–2.69)	1.21 (0.75–1.94)	1.59 (0.98–2.57)	0.81 (0.54–1.20)
Two or more sexual partners	1.27 (0.77–2.11)	1.50 (0.90–2.51)	0.66 (0.40–1.08)	1.26 (0.83–1.91)
Ever had sex without the use of birth control	0.95 (0.61–1.48)	1.09 (0.70–1.70)	1.78 (1.13–2.79)	0.81 (0.60–1.19)
One or more pregnancies	0.94 (0.59–1.49)	1.00 (0.62–1.61)	1.00 (0.63–1.59)	1.01 (0.67–1.52)
Relationship type (ref: casual)
Nonexclusive dating	1.16 (0.41–3.30)	1.25 (0.47–3.33)	0.92 (0.35–2.46)	1.47 (0.57–3.81)
Long-distance	1.05 (0.38–2.84)	0.61 (0.23–1.61)	0.72 (0.28–1.82)	2.18 (0.88–5.39)
Exclusive dating	0.70 (0.29–1.67)	0.48 (0.21–1.08)	0.34 (0.15–0.76)	1.72 (0.78–3.77)
Cohabiting	0.48 (0.18–1.30)	0.28 (0.11–0.73)	0.30 (0.12–0.75)	2.28 (0.95–5.43)
Married/engaged	1.09 (0.42–2.84)	0.54 (0.21–1.38)	0.31 (0.12–0.78)	1.42 (0.59–3.43)

High school GPA was significantly associated with all four distress measures, that is, women with higher GPAs were less likely to experience depression, stress, and loneliness but more likely to have high self-esteem. Additionally, women with early ages at first sex were more likely to experience depression, and women who had sex without the use of birth control were more likely to experience loneliness. Of note, as relationship seriousness increased, loneliness decreased.

Table 3 presents the effect of each measure of baseline PD on subsequent (journal) IPV controlling for sociodemographic characteristics, adolescent experiences related to sex and pregnancy, and relationship type net of past IPV. Depression, stress, and loneliness were positively associated with IPV, while higher self-esteem was negatively associated with IPV (i.e., women with high self-esteem were less likely to experience IPV).

Table 3.

Multinomial Logistic Regression Results (Odds Ratios) of the Effect of Baseline Psychological Distress on Subsequent Violence (N = 23,600 Weeks)

	Depression		Stress		Loneliness		High self-esteem
	Only psychological	Any physical	Only psychological	Any physical	Only psychological	Any physical	Only psychological	Any physical
Psychological distress	1.37 (1.21–1.56)	1.04 (0.76–1.42)	1.44 (1.28–1.62)	2.01 (1.52–2.65)	1.73 (1.54–1.94)	1.26 (0.94–1.68)	0.78 (0.70–0.86)	0.83 (0.64–1.07)
Past violence (ref: none)
Only psychological	5.06 (4.48–5.72)	6.22 (4.74–8.17)	4.95 (4.38–5.59)	6.64 (5.07–8.70)	4.91 (4.35–5.55)	5.92 (4.51–7.76)	4.84 (4.29–5.47)	6.55 (4.99–8.59)
Any physical	4.41 (3.26–5.98)	15.09 (9.33–24.43)	4.40 (3.25–5.97)	13.99 (8.76–22.36)	4.42 (3.26–5.99)	14.33 (8.99–22.85)	4.36 (3.22–5.90)	15.04 (9.46–23.90)
Sociodemographics
19 (ref: 18)	0.96 (0.86–1.07)	0.69 (0.53–0.91)	0.94 (0.84–1.05)	0.70 (0.54–0.92)	0.93 (0.83–1.04)	0.69 (0.53–0.90)	0.96 (0.86–1.07)	0.70 (0.53–0.91)
20 (ref: 18)	0.98 (0.81–1.19)	0.51 (0.30–0.87)	0.99 (0.82–1.20)	0.58 (0.34–0.99)	0.97 (0.80–1.17)	0.52 (0.30–0.88)	0.98 (0.81–1.19)	0.52 (0.30–0.88)
African American (ref: non-African American)	0.73 (0.63–0.85)	1.81 (1.30–2.53)	0.75 (0.65–0.87)	1.77 (1.27–2.46)	0.69 (0.59–0.81)	1.74 (1.24–2.44)	0.77 (0.66–0.90)	1.83 (1.31–2.55)
Highly religious	1.08 (0.96–1.20)	0.89 (0.67–1.19)	1.09 (0.97–1.21)	0.83 (0.63–1.10)	1.13 (1.01–1.26)	0.90 (0.68–1.20)	1.12 (1.00–1.25)	0.91 (0.68–1.20)
Childhood disadvantage	0.96 (0.91–1.02)	0.85 (0.74–0.98)	0.96 (0.91–1.02)	0.85 (0.74–0.97)	0.99 (0.93–1.04)	0.86 (0.75–0.99)	0.96 (0.91–1.02)	0.85 (0.74–0.97)
High school GPA	0.72 (0.66–0.78)	0.68 (0.55–0.83)	0.73 (0.67–0.79)	0.69 (0.57–0.85)	0.75 (0.69–0.82)	0.69 (0.56–0.85)	0.73 (0.67–0.80)	0.69 (0.56–0.84)
Receiving public assistance	1.13 (0.97–1.31)	1.73 (1.23–2.45)	1.14 (0.98–1.32)	1.14 (0.98–1.32)	1.07 (0.92–1.24)	1.70 (1.20–2.40)	1.08 (0.93–1.26)	1.72 (1.22–2.43)
Adolescent experiences related to sex and pregnancy
Age at first sex ≤16 years	1.00 (0.87–1.15)	1.23 (0.85–1.76)	1.01 (0.88–1.16)	1.23 (0.86–1.77)	1.05 (0.92–1.21)	1.23 (0.86–1.77)	1.03 (0.90–1.19)	1.23 (0.86–1.77)
Two or more sexual partners	1.01 (0.88–1.16)	2.20 (1.47–3.28)	1.04 (0.90–1.19)	2.07 (1.40–3.06)	1.05 (0.91–1.21)	2.19 (1.47–3.27)	1.09 (0.95–1.25)	2.25 (1.51–3.35)
Ever had sex without the use of birth control	1.21 (1.07–1.37)	1.09 (0.81–1.46)	1.19 (1.05–1.35)	1.09 (0.81–1.46)	1.15 (1.02–1.31)	1.07 (0.80–1.44)	1.18 (1.04–1.34)	1.09 (0.81–1.46)
One or more pregnancies	0.81 (0.70–0.95)	0.43 (0.29–0.62)	0.82 (0.71–0.95)	0.41 (0.28–0.60)	0.81 (0.70–0.95)	0.43 (0.30–0.62)	0.82 (0.71–0.96)	0.42 (0.29–0.61)
Relationship type (ref: casual)
Nonexclusive dating	1.24 (0.79–1.96)	1.73 (0.46–6.51)	1.25 (0.80–1.98)	1.79 (0.48–6.72)	1.22 (0.78–1.93)	1.73 (0.46–6.49)	1.26 (0.80–1.99)	1.76 (0.47–6.60)
Long-distance	2.35 (1.68–3.28)	1.47 (0.49–4.40)	2.43 (1.74–3.40)	1.57 (0.52–4.71)	2.37 (1.69–3.32)	1.48 (0.50–4.44)	2.36 (1.69–3.31)	1.48 (0.49–4.43)
Exclusive dating	1.79 (1.29–2.48)	3.31 (1.20–9.10)	1.87 (1.34–2.59)	3.56 (1.29–9.82)	1.80 (1.30–2.50)	3.35 (1.22–9.22)	1.81 (1.30–2.51)	3.38 (1.23–9.30)
Cohabiting	3.88 (2.80–5.39)	8.48 (3.09–23.28)	4.11 (2.96–5.71)	9.74 (3.54–26.82)	4.03 (2.90–5.60)	8.63 (3.14–23.71)	3.87 (2.79–5.38)	8.50 (3.09–23.36)
Married/engaged	1.77 (1.27–2.47)	3.44 (1.23–9.61)	1.85 (1.32–2.59)	3.83 (1.37–10.71)	1.85 (1.32–2.59)	3.49 (1.25–9.77)	1.76 (1.26–2.46)	3.44 (1.23–9.63)

The reference category for violence is none.

Women who reported depression, stress, loneliness, and low self-esteem at baseline were more likely to subsequently experience psychological violence only relative to no violence compared with women who did not have a history of PD. Women who reported stress were also more likely to experience any physical violence relative to no violence compared with women who did not have a history of stress. Depression, loneliness, and self-esteem, however, were not significantly associated with the risk of any physical violence net of individual and relationship characteristics.

The associations between individual and relationship characteristics and IPV were similar net of each PD measure as well as past IPV. African American race had a protective role against experiencing only psychological violence, but was correlated with an increased risk of any physical violence. Higher high school GPA decreased the risk of experiencing either type of IPV net of each PD measure. Women receiving public assistance at age 18 or 19 were more likely to experience any physical violence relative to no violence.

For adolescent experiences related to sex and pregnancy, having two or more sexual partners during adolescence was significantly associated with an increased risk of physical violence. Women who had ever had sex without the use of birth control during adolescence were found to be at an increased risk of psychological violence. In contrast, women who had pregnancies during adolescence were less likely to experience either type of violence. Generally the more serious a relationship was, the higher the risk was for experiencing violence, particularly psychological violence.

Discussion

The literature on IPV^3–5,9 has established a link between IPV and subsequent PD; recent research has begun to look at how PD may precipitate IPV.⁸ In this study, we examined the PD consequences of IPV, as well as the IPV consequences of PD. The results demonstrated that the IPV–PD relationships are bidirectional.

Women who had experienced past IPV were more likely to report depression, stress, loneliness, and low self-esteem at baseline. Conversely, those who were psychologically distressed at baseline had an increased risk of subsequent IPV, particularly psychological violence even after controlling for past IPV, sociodemographic characteristics, adolescent experiences related to sex and pregnancy, and relationship type. Stress was the only distress measure that was significantly associated with both psychological and physical IPV net of controls. In addition, high school GPA had a consistent protective effect on both IPV and distress. The incidence of IPV increased with relationship seriousness.

This study builds upon previous research by examining longitudinal IPV experiences among emerging adult women. Furthermore, this study focuses on multiple indicators of PD (depression, stress, loneliness, and low self-esteem) to provide a better understanding of the varied impact of distress on emerging adults.

Health care providers are often the first encounter for women who experience IPV, and thus interactions with primary care or emergency department settings are crucial opportunities for IPV intervention. Patients who experienced IPV are more likely to utilize health care services than those who did not experience violence.²⁷ However, IPV screening and identification outcomes remain inconsistent and low. Time constraints, lack of standard protocols, and personal perceptions toward IPV screening deter providers from addressing IPV.²⁷ Nicolaidis et al. have found that women who received health care services as a result of IPV felt that health care providers did not adequately explore the complexity of their experiences.²⁸ As IPV and PD are closely related, screening for IPV and distress would prevent further health complications.

Limitations

There are several limitations to this study. First, due to the weekly frequency (and the breadth of topics to be addressed), RDSL respondents were asked a limited number of IPV questions. Though RDSL included only a few measures of psychological and physical violence, the strength of the RDSL measurement strategy was that it allowed for a continuous (weekly) recording of violence for every relationship a respondent experienced during the study period. In addition, RDSL collected only IPV victimization data from women but not their partners. Although IPV is perpetrated by both men and women, gender-based violence against women remains more prevalent and harmful.²⁹ Further, RDSL did not measure sexual violence on a weekly basis. Second, PD was only measured at baseline. Despite this, the longitudinal assessments of violent experiences inform the relationship between a past history of IPV and PD in addition to the effects of PD on subsequent IPV.

Lastly, although the RDSL sample was randomly selected and population-representative, it consisted of women residing in a single Michigan county, which may decrease generalizability. However, focusing on a single county reduced variation in other characteristics that were not of interest here (e.g., local availability of mental health services). We also did not expect that the underlying processes between IPV and PD will vary across regions. Further research on the dynamic role between IPV and PD among a more regionally diverse sample would provide additional insights.

Conclusions

The increased incidence of IPV and PD during the already challenging emergence into adulthood poses a risk to normative development, emphasizing the need for standardized intervention protocols that address both IPV and PD.

Author Disclosure Statement

No competing financial interests exist.