The Centers for Disease Control (CDC 2007) has defined IPV as physical, sexual, or psychological abuse by a spouse, ex-spouse, boyfriend, girlfriend, ex-boyfriend, ex-girlfriend, or date. Research on intimate partner violence (IPV) has greatly increased over the past several decades. Most of the research has been dedicated to domestic violence with a predominant focus on adult women in abusive relationships (Cascardi and O’Leary 1992; Stein and Kennedy 2001). Recently, interest has broadened to IPV in adolescent relationships. A report by the Bureau of Justice Statistics of the Department of Justice (2001) revealed that 22% of female homicide victims between the ages of 16 and 19 were killed by intimate partners during the years of 1993 to 1999. Further, as many as one in five girls will be involved in an abusive relationship during adolescence (Silverman et al. 2001). Adolescents who are mothers and poor appear to be at an elevated risk of IPV (Leadbeater et al. 2001; Rennison 2001).

Despite statistics indicating that IPV in adolescents is a widespread crisis with potentially fatal consequences, there is a paucity of research in this area. Researchers have started to understand the risk factors, characteristics, and mental health correlates of the phenomenon; however, there are still many gaps as well as contradictory findings in the literature. One group of adolescents who may be particularly vulnerable to IPV is pregnant and post-partum adolescent mothers (Gielen et al. 1994; Parker et al. 1993). However, little is known about the psychological correlates of IPV for this group. Moreover, few investigators have compared adolescent and young adult mothers in the same study.

Pregnancy

Pregnant females have been identified as a vulnerable group that experiences high rates of IPV (Burch and Gallup 2004; Gelles 1988). Martin et al. (2004) linked pregnancy onset to increases in psychological and sexual aggression. Parker et al. (1993) found a significantly higher rate of IPV during pregnancy among adolescents compared to adult women. In their study of 545 Medicaid-eligible females, Covington et al. (2001) found prenatal violence (by partners, relatives, and acquaintances) to be more common among adolescents (16.1%) than adult women (11.6%). Further, prevalence rates of IPV among adolescents have been found to increase during the postpartum period. In a study of 275 adult and adolescent females, 19% reported IPV during pregnancy while 25% reported postpartum IPV (Gielen et al. 1994). The rate of IPV toward adolescent mothers by male partners has been reported to be as high as 41% (Leadbeater et al. 2001).

Risk for IPV among adolescent mothers may also be exacerbated by the tendency for many of these pregnancies to have been unplanned. Indeed, research has shown that females with unplanned pregnancies have four times the odds of experiencing IPV than females with planned pregnancies (Gazmararian et al. 1995). In addition, low income adolescents experiencing IPV are more likely to have rapid, repeat pregnancies (Jacoby et al. 1999). Thus, as a result of their young age, unplanned pregnancies, and increased postpartum stress, the existing evidence suggests that adolescent mothers may be at an increased risk of partner violence relative to young adult mothers. Despite their increased vulnerability, however, only a small number of studies of adolescents have investigated pregnant and parenting adolescents (Gielen et al. 1994; Parker et al. 1993; Silverman et al. 2004); and most of these studies have not examined the psychological correlates of IPV among this population.

Developmental Stage and IPV

Although an abundance of research exists on adult IPV, these findings cannot simply be generalized to adolescent romantic relationships. Researchers have provided evidence supporting a variety of differences between adolescent and adult relationships. First, research has found that not all who use physical force in dating relationships become batterers once married and not all spouses who batter used force in dating relationships (Follingstad et al. 1999). Second, researchers have discovered that adolescent mothers experience IPV at a significantly higher rate than adult mothers (Gielen et al. 1994; Parker et al. 1993). Third, adolescents tend to lack dating experiences as compared to young adults (Montgomery 2005) and, therefore, may accept physical and sexual aggression as normal in dating and partner relationships (Hamberger and Ambuel 1998). In addition, girls may pursue a relationship before they are ready in order to be accepted by their peers (Smith and Donnelly 2001). As a result, they may not perceive physically aggressive behavior as wrong or destructive and may not take action to protect themselves from abusive acts.

Definitions of adolescence and young adulthood vary across different studies as well as cross-culturally. Arnett and Taber (1994) argued that the age at which one is considered a young adult is socially defined and varies among cultures. Further, when indicators of the achievement of adult status (such as pregnancy or marriage) occur in adolescence, the exact age of transition becomes even more difficult to determine. Arnett (2001) has argued that age 20 can be one possible transition point into adulthood, which is consistent with the majority of empirical research that includes youth under the age of 20 in the adolescent category. For purposes of the present study, we define “adolescent mothers” as those under age 20 in order to facilitate a comparison of our findings on adolescents to the existing literature.

The literature on IPV in adult mothers typically does not distinguish between young adults and older adult mothers, despite the statistics indicating that younger mothers are less likely to be married at the time of the baby’s birth (Martin et al. 2006). Given that unmarried individuals have fewer personal or financial resources (Waite 1995), and unmarried women report higher levels of IPV than dating women (Hamberger and Ambuel 1998), it is likely that unmarried young adults are at relatively higher risk than older adults. Moreover, as indicated above, pregnancy and parenting may exacerbate this risk. In light of this previous research, in addition to adolescents, the present study also examines “young adult mothers” who were ages 20 or 21, as they may represent a subgroup of mothers who have an increased risk for IPV relative to their older adult counterparts.

IPV and Mental Health

Research on adult romantic relationships has firmly linked symptoms of anxiety and depression to IPV (Cascardi and O’Leary 1992; Hamberger et al. 1993; Stein and Kennedy 2001). Among adolescents, researchers have examined the relationship between IPV and the broad concepts of self-esteem and emotional well-being but have less often specifically investigated anxiety and depression. The studies that have addressed anxiety and depression have indicated a trend towards elevated levels of anxiety and depression with some contradictory results. Sussex and Corcoran (2005) reported that pregnant and parenting adolescents experiencing IPV had significantly higher rates of depression compared to non-abused teen mothers. Even after the cessation of violence, elevated rates of depression among pregnant and parenting adolescents continued indicating the strong impact of IPV on depressed mood. Studies conducted with adolescents (who were not pregnant or parenting) reported similar results. Kilpatrick et al. (2003) gathered data through phone interviews and discovered that IPV increased the risk of depressive and anxious symptoms among adolescent females. Partly supporting these results, Callahan et al. (2003) reported that high school girls experiencing dating violence displayed posttraumatic stress and dissociation but not depression. Conversely, Boney-McCoy and Finkelhor (1996) found an increased risk of depression after experiencing violence among 10–16 year old females. However, these results are not specific to IPV and include all types of violence. Additional research is needed to isolate the impact of IPV using a national sample to support the trend of elevated rates of anxiety and depression associated with IPV, especially in vulnerable populations such as young mothers.

IPV Among Latinos

Among adults, research has demonstrated that Latino women are as much as 50% more likely to experience domestic violence than Caucasians and African Americans (Kessler et al. 2001). However, other studies have found domestic violence rates to be almost equivalent across ethnic groups (e.g., Bassuk et al. 2006). Torres (1991) speculated that these discrepant results may be due, in part, to the tendency of Latino women to underreport domestic violence because they want to protect their partners. McGee (1997) performed a meta-analysis on literature addressing cultural values and domestic violence and concluded that Latino women may tolerate and acknowledge IPV as part of their subordinate role in a machismo culture. Studies of IPV among Latino adolescents have yielded mixed findings. While research conducted by Coker et al. (2000) found little difference in IPV rates between Caucasian and Latino adolescents in South Carolina (9% and 8% respectively), other studies have found slightly higher rates among Latino adolescents. Specifically, the National Youth Risk Behavior Study (Eaton et al. 2006) revealed that approximately one in ten (9.9%) Latino youth experienced an abusive relationship relative to 8.2% of Caucasian youth. Even greater discrepancies in rates of IPV among Latinos (11.31%) as compared to Caucasians (7.43%) were reported by Howard and Wang (2003). Finally, other research has found that Latino adolescents and adults were as much as 50% more likely to experience IPV than Caucasians (Kessler et al. 2001). Reasons for these inconsistencies may be due to the variations in prevalence based on geographic region or due to limited measurement of IPV using a single-item (Eaton et al. 2006; Howard and Wang 2003).

Given the potential for elevated IPV risk among Latino adolescent and adult women, which may be exacerbated by the stress of pregnancy and parenting, researchers have recognized the importance of investigating Latino mothers’ experiences (Nadeem et al. 2006). However, few existing studies have reported on the prevalence of IPV among Latino adolescent mothers. The current study will address the limitations of previous research by reporting IPV prevalence among Latino adolescents and young adults using a multidimensional measure to assess IPV. Further, this study will provide prevalence rates using a large, national sample of adolescents and young adults, thus improving generalizability to the Latino community.

Marriage and IPV

A final risk factor for IPV may be marital status. Research suggests that unmarried romantic relationships may be more violent than married relationships. Hamberger and Ambuel (1998) found violence among dating couples to be more common than among married couples. Johnson et al. (2003) discovered that boyfriends, as opposed to husbands, were the most common perpetrators. Across ethnic groups, lower rates of IPV among married couples have been consistently reported (Flake and Forste 2006; Vest et al. 2002). Thus, the evidence suggests that non-married women might be at higher risk for IPV relative to married women. However, it is unclear whether these findings generalize to unmarried parenting adolescents and young adults.

Summary

IPV among adolescents has received minimal research attention, but is prevalent and potentially dangerous. While it is important to investigate adolescent IPV as a unique phenomenon, it is helpful to compare the experiences of adolescent with young adult mothers to discover whether extensive research on adult IPV can be applied to adolescent mothers. Depression and anxiety have been firmly linked to abusive relationships among adults (Hamberger et al. 1993; Stein and Kennedy 2001); and higher rates of IPV have been found among Latino adolescents and adults (Howard and Wang 2003; Kessler et al. 2001). Since IPV rates have been shown to increase in the postpartum period (Gelles 1988), especially among adolescents (Gielen et al. 1994), and decrease among married couples (Hamberger and Ambuel 1998), vulnerable unwed adolescents may be particularly at risk for IPV and poor mental health after giving birth.

Current Study

Despite increased research on adolescent IPV, many unanswered questions remain and there is a preponderance of contradictory results. The clearest finding of the research is that adolescent dating violence is a significant problem which has been relatively unaddressed, particularly among young mothers. Utilizing the CDC definition of IPV, the current study attempts to fill gaps in previous research by providing information about adolescent and young adult mothers experiencing IPV.

Using a sample at high risk for IPV, the following hypotheses were examined: (1) IPV rates will be higher among the adolescent mothers as compared to the young adult mothers. In addition, rates of depression and anxiety among adolescents and young adults will be compared on an exploratory basis because earlier studies have not compared mental health correlates between young adult and adolescent mothers; (2) IPV will be positively associated with symptoms of depression and anxiety; (3) Mothers with Latino heritage, as opposed to those without Latino heritage, will have higher rates of IPV; and (4) Married mothers will report lower levels of IPV.

Method

Procedure

Data were obtained from the Fragile Families and Child Wellbeing Study (McLanahan et al. 1998, unpublished), a nationally representative survey of births to parents in 20 American cities. Cities had a population of at least 200,000 with varied welfare policies, child support enforcement laws, and labor market conditions. Mothers were recruited in hospitals and gave their first interview within 48 h of the birth of their child. Baseline data collection occurred between June, 1999 and March, 2002. Follow-up interviews were conducted approximately 1 year later either by phone (one-third) or in person. All mothers and fathers were interviewed separately in Spanish or English, and were compensated with $20 for each interview. Data gathered from mothers at the 1-year follow-up interview were used for the current study.

Participants

The present sample consisted of 672 mothers between the ages of 14 and 21 years (M = 19.48 years, SD = 1.14) who had given birth 1 year prior to the interview. The majority of participants reported their race to be African American (59%), with the remaining participants reporting their race as Caucasian (24%), American Indian (2.1%), Asian American (1.5%) or Other (13.4%). In the total sample, 24.5% described themselves as Hispanic or Latino, considering their descent to be Mexican (41.3%), Puerto Rican (19.8%), Cuban (1.2%), Central American/Caribbean (3.5%), South American (1.8%), Spanish (1.2%) or Other (21.5%). Among all the mothers, 10.5% were married at the time of the follow-up interview. The majority (65.8%) of the women had only one child, while most (26.1%) of the other mothers had two children. A large percentage (66.3%) of the women was not attending school or other training classes but most (75.6%) women had worked in the last 12 months. Of the women who had worked, many (33.9%) of them earned a yearly income of $10,000 and below or made (40.1%) between $10,000 and $40,000. In addition, 34.7% of the mothers had received welfare or TANF within the preceding 12 months.

In order to compare young adult mothers to adolescent mothers, two groups were created. The “adolescent mother” group included participants between the ages of 14 and 19 (n = 282) and the “young adult mother” group included participants between the ages of 20 and 21 (n = 390). Table 1 provides a comparison of the two sub-samples on key demographic and descriptive variables.

Table 1 Descriptive information for adolescents and young adult mothers
Full size table

Measures

Depressive Symptoms

Questions that assessed levels of depression were derived from the Composite International Diagnostic Interview-Short Form (CIDI-SF; Kessler et al. 1998). The CIDI-SF questions are based on the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV; APA 1994). The CIDI-SF is a standardized instrument used for assessing mental disorders that can be used in cross-cultural research studies.

Participants were asked whether they had feelings of depression or sadness in the previous year that lasted for 2 weeks or more and whether there was a time lasting 2 weeks or longer when they lost interest in activities which usually gave them pleasure. Each question yielded a “yes” (1) or “no” (0) answer. The two items are criteria for diagnosing Major Depressive Disorder according to the DSM-IV (APA 1994) and were used as screening items, only answering additional questions about their symptoms if they endorsed at least one item. Therefore, we only used these two items because they were answered by all participants. A dichotomous depression score was calculated for each participant indicating whether they reported symptoms of depression (1 = yes, 0 = no).

Anxiety Symptoms

Presence of anxiety was assessed using two questions from the CIDI-SF (Kessler et al. 1998). Participants reported whether they had felt excessive worry or anxiety about more than one thing and for more days than not or if they had difficulty controlling their concerns or anxiety during the past 6 months. Each question yielded a “yes” (1) or “no” (0) answer. Participants only answered more questions regarding anxiety symptoms if they endorsed either of these items. Therefore, we only used these two items because they were answered by all participants. A dichotomous anxiety score was calculated for each participant indicating whether they reported symptoms of anxiety (1 = yes, 0 = no).

IPV

IPV was measured from questions that were derived from questionnaires created for the Effects of Violence on Work and Family Project (Lloyd 1997). These measures were based on the Conflict Tactics Scale (Straus 1979) but included additional questions which incorporated a broader range of physical and emotional means by which men coercively control women. Participants were asked questions such as “He slaps or kicks you,” “He tries to prevent you from going to work or school,” “He insults or criticizes you or your ideas” and “He hits you with his fist or an object that could hurt you.” Each question was responded on a three-point Likert scale (3 = often, 2 = sometimes, 1 = never). Items were summed to create a composite score for each participant, with higher scores indicating higher levels of IPV. The mean score for the total sample was 8.3 for the total sample, with identical means for each sub-sample (range = 7 to 21). The scale demonstrated reasonable internal consistency in the current sample (α = 0.60).

Demographics

Participants were asked a variety of demographic questions such as the number of biological children they have and whether they worked or received welfare during the previous year. In addition, to assess Latino heritage and marital status, the participants were asked if they were of Hispanic or Latino origin or descent and if they were currently married.

Results

Logistic and linear regressions, correlations, and descriptive analyses were conducted on the data. Correlations coefficients for the major study variables are found in Table 2. The adolescent and young adult data were analyzed separately and then comparison analyses were conducted between the two groups. All continuous variables were tested using centered data.

Table 2 Correlation coefficients of key variables for adolescent and young adult mothers
Full size table

Comparison of Adolescent and Young Adult Mothers

The first hypothesis that adolescent mothers would report higher levels of IPV than young adult mothers (adolescents: μ = 8.27, SD = 2.28; young adults: μ = 8.27, SD = 2.23) was not supported. An exploratory comparison of mental health symptoms between adolescents and young adults led to chi-square analyses which revealed significant differences for reports of anxiety (p = 0.00, χ 2 = 11.61) and reports of depression (p = 0.00, χ 2 = 8.69) such that young adults reported symptoms with greater frequency. Among the young adults, more symptoms of anxiety were reported than symptoms of depression. Exploratory comparisons revealed that anxiety was reported with much greater frequency among the young adults (32%) than the adolescents (19%). Similarly, depression was reported more frequently among the young adults (25%) than the adolescents (13%). Not surprisingly, the same was true of marriage (young adults 12%, adolescents 8%).

We considered the possibility that women who lived with their partners may have experienced higher rates of IPV because close proximity to their partner may result in more opportunities for abuse and violence to occur. To explore this possibility, a post-hoc analysis examined differences in IPV based on cohabitation status. Among the young adults, approximately equal numbers lived with their partners 201 (49%) and separate from their partners 195 (51%). Similarly in the adolescent sample, 129 (40%) participants lived with their partners and 190 (60%) adolescents did not live with their partners. The difference in rates of IPV between those who were and were not cohabitating was not statistically significant.

Analysis of Variables with Adolescent Mothers

To test the second hypothesis, that IPV would be significantly positively associated with symptoms of depression and anxiety among adolescents, logistic regressions were conducted with IPV as the independent variable and anxiety and depression as dependent variables (see Table 3). Contrary to the hypothesis, IPV was not associated with depression or anxiety among adolescents. This may be due, in part, to the low level of symptoms reported by adolescents. To explore the third and fourth hypotheses, a linear regression using Latino heritage and marriage status as independent variables and IPV as the dependent variable (see Table 4) was conducted. The third hypothesis, that women with Latino heritage, as opposed to those without Latino heritage, would have higher rates of IPV, was not supported. However, there was a significant negative correlation between Latino heritage and marriage (r = −0.17). There was also no support for the fourth hypothesis that married adolescents would report lower levels of IPV.

Table 3 Summary of logistic regression results with IPV as independent variable
Full size table
Table 4 Linear regression results with IPV as dependent variable
Full size table

Analysis of Variables with Young Adult Mothers

Next, the hypotheses 2, 3 and 4 were tested with the young adult mothers’ data using parallel analyses to those conducted with the adolescent mothers’ data. When examining the second hypothesis, that IPV would be significantly positively associated with symptoms of depression and anxiety among young adults, logistic regressions (see Table 3) revealed significant positive associations between IPV and depressive symptoms (B = 0.14, p = 0.01) as well as between IPV and anxiety symptoms (B = 0.13, p = 0.01). The linear regression to test hypothesis 3, that young adults with Latino heritage will have higher rates of IPV, was not supported. However, the fourth hypothesis that married young adults would report lower levels of IPV was supported (see Table 4).

Discussion

Young mothers, particularly adolescent mothers, appear to be at high risk for IPV (Leadbeater et al. 2001; Parker et al. 1993) and consequently psychological distress (Kilpatrick et al. 2003). However, there is a paucity of research on the mental health consequences of IPV among adolescents as well as the differences between IPV among adolescent and young adult mothers. IPV can be life-threatening which makes the current study critical for informing interventions which can improve the welfare of young mothers and their families.

The results partially supported the proposed hypotheses. Consistent with previous research, the young adult mothers in our sample who were experiencing IPV reported more symptoms of depression and anxiety than their counterparts who were not experiencing IPV (e.g., Stein and Kennedy 2001). However, these associations were not significant in the adolescent mother sub-sample. Further, adolescent mothers did not report higher levels of IPV than young adult mothers. Despite similar rates of IPV being reported across the age groups, fewer symptoms of anxiety and depression were reported among the adolescents as compared to the young adults. Previous research on mental health among adolescents experiencing IPV has reported mixed findings. Therefore, the results are consistent with those that did not find an association of IPV with depression or anxiety (Callahan et al. 2003). Differing from our hypothesis, reports of IPV were not higher among Latino adolescents or young adults as compared to their non-Latino peers. However, consistent with our hypothesis, marriage was associated with lower levels of reported IPV in the young adult sample. While marriage and IPV were not associated in the adolescent sample, marriage was related to lower levels of anxiety in the group.

IPV was related to symptoms of depression and anxiety among young adult mothers, but not adolescent mothers, which partly supports the hypothesis. However, these results do support the theory that adolescent IPV is a unique phenomenon (e.g., Follingstad et al. 1999); and research on adult IPV and mental health correlates should be replicated using adolescent populations. Adolescents are inexperienced in relationships and have been found to interpret abuse as love (Levy 1990) as well as to consider violence as normal in relationships (Hamberger and Ambuel 1998). Possibly, adolescent females deal with their abusive experiences by hitting back instead of becoming depressed or anxious. Adolescents might be expressing their response to IPV through aggression toward their partners and, thereby, protecting their mental health. Adolescent perpetrators and victims have been found to have equally violent profiles with both partners perpetrating and sustaining physical, sexual and emotional aggression (Avery-Leaf et al. 1997). Further, the balance of power between genders in romantic relationships appears to be more equal among adolescents than young adults (O’Leary and Slep 2003) with research starting to investigate female aggression within these relationships (Holtzworth-Munroe 2005). This indicates a possibility that there is a coercive and violent mutual relationship during adolescence. Moreover, the sample in the current study represents a unique group, adolescent mothers, which exist within the adolescent population. The results could be indicative of a distinct pattern of IPV and mental health in the parenting adolescent population.

Additional unique characteristics of adolescence may account for the differences in mental health symptoms between sub-groups. During adolescence, acceptance by peers is of the utmost importance. Girls may feel accepted by peers because they are in relationships (Smith and Donnelly 2001). If girls are developing depressed or anxious symptoms after abuse, the boost to self-esteem that they receive from peers who support their relationship may temporarily suppress the mental health impact of the abuse. However, as people age, peer acceptance becomes less important, so the benefits which come from it may also be reduced. The protective quality of peer influence may disappear after adolescence leaving young adults more vulnerable to depression or anxiety. Therefore, during adolescence the abuse may set in motion a series of events or vulnerabilities that do not impact psychological well-being until young adulthood. The unique characteristics of adolescent relationships may be some of the reasons that more symptoms of depression and anxiety were reported in the young adults as opposed to the adolescents in our sample. While the results from the current study indicate less severe mental health consequences of IPV for adolescent mothers as compared to young adult mothers, there may be a negative long-term impact on adolescent mothers’ ability to form and maintain healthy relationships and a dangerous short-term impact of IPV on the physical health of adolescent mothers (Silverman et al. 2004). Moreover, parents modeling a violent relationship may lead to an intergenerational transmission of violence to their children (Fantuzzo and Mohr 1999; Gagne et al. 2005; Maker et al. 1998). Longitudinal studies are needed to examine these issues.

As opposed to temporarily subdued symptoms of depression or anxiety, adolescents may have had a mental health response to IPV that did not present as depression or general anxiety. These adolescents may have coped with their experience through dissociation and numbness, which was not captured in the current study. This explanation is consistent with research reported by Callahan et al. (2003), but inconsistent with other investigators (Boney-McCoy and Finkelhor 1996; Kilpatrick et al. 2003). Possibly certain adolescents have a response similar to PTSD symptoms while others respond more with depressed or anxious symptoms. Future research that examines unique symptoms of other anxiety disorders such as PTSD will be helpful in understanding the current results.

Latino heritage was not linked to IPV among the adolescents or young adults in our sample. Acculturation levels may have been suppressing a statistical association between IPV and Latino heritage. Latino women have tended to underreport IPV in past research (Torres 1991). The present sample might be less acculturated and may therefore endorse traditional beliefs about IPV (Torres) and, thus, underreport coercive behaviors in the survey. However, previous research on abusive relationships among Latino youth has been inconsistent when comparing acculturated to less acculturated participants (Ackard and Neumark-Sztainer 2002; Coker et al. 2000). In the current study, the mean rates of IPV in the Latino group were slightly elevated, but these differences were not statistically significant. Unfortunately, acculturation level was not able to be assessed in the present study, so we can only conjecture about its impact on IPVFootnote 1.

Rennison and Planty (2003) suggest that IPV rates are similar across ethnicities after SES status is accounted for, however, this cannot explain the results in the current sample of low SES females. Yoshioka et al. (2003) reported that abused Latino adult women most often turn to family members for support. And, while many Latino women in their sample sought assistance from non-kin members, the participants were a nonrandom sample of women who had accessed domestic violence services. Our sample was not chosen based on IPV experiences. Therefore, perhaps they represent Latino women who are not reporting abusive experiences to people outside of the kinship network. Conversely, as Edelson et al. (2007) found, the rates might be the same across ethnic groups but the effect of IPV may differ. More research is needed to examine rates and outcomes of IPV in random samples of Latino and non-Latino young adult and adolescent mothers.

Supporting previous research (Hamberger and Ambuel 1998; Parker et al. 1993), this study found that young adults who were married had lower rates of IPV. However, in the adolescent sample, IPV was not related to marriage but marriage was correlated with lower rates of anxiety. Adolescent mothers may receive greater judgment from others or experience more financial stress and uncertainty than young adult mothers, resulting in anxiety. Anxiety symptoms may be reduced by having a spouse. However, this study was cross-sectional and the causal direction cannot be inferred. Future longitudinal research should explore and compare adolescent mothers who are married to those who are not married and that impact on mental health.

There were several limitations to the current study. First, depression and anxiety were measured as categorical variables instead of continuous variables. As a result, variation in the degree of psychological distress was not captured. In addition, there were a limited number of questions used to detect symptoms of depression and anxiety. Mental health correlates with IPV among adolescents may have been discovered if continuous variables and more extensive measures had been used. However, the variables used in the study are required for a DSM-IV diagnosis; therefore these results likely captured those whose symptoms were severe enough to meet diagnostic criteria. Nonetheless, participants who had symptoms but were sub-threshold for diagnosis may have been missed.

Second, previous research has demonstrated that adolescents may interpret abuse differently from young adults (e.g., Levy 1990) which indicates that they may describe the abuse differently than was presented in the questionnaire. Further, since adolescents may perceive abuse to be love, they may refrain from endorsing items in the questionnaire which appear negative when referring to IPV. A disconnect may exist between their personal “positive” experiences and the “negative” descriptions in the questionnaire even though both are describing the same objective event. It appears that adult IPV measures may not accurately describe adolescents’ experiences and a modified measure is needed to specifically capture adolescents’ experiences of IPV. Further, our sample of adolescents includes youth from age 14 through age 19, with the majority of the group being ages 18–19. Therefore, it is possible that the younger adolescents might have different perceptions and experiences due to lower levels of maturity. Unfortunately, due to sample size limitations, we were unable to examine the younger adolescents separately from older ones, but this would be an important direction for future research. Similarly, future research could examine young adults who are in their mid-20s to explore whether the higher levels of reported depressive and anxiety symptoms continue further into adulthood or abate as these mothers get older.

Third, as noted earlier, the present study did not assess acculturation level which may be an important factor in the association between IPV and Latino heritage. Unique cultural factors, such as “machismo” or “respeto”, may have impacted Latino female’s experiences of IPV (Perilla 1999; Perilla et al. 1994). Latino adolescents and young adults in the current study may have been impacted by their acculturation level in deciding whether to report abuse or perceive violent acts negatively. In addition, a paucity of research also exists in the area of acculturation of Latino youth and its impact on relationships and IPV. Future research could investigate the extent to which Latino youth who immigrated to the USA at a young age endorse traditional ideas of aggression in relationships.

In addition to investigating the experience of abusive relationships for Latino adolescents, there are many other directions for future research in the area of adolescent IPV. Research would benefit from an in-depth analysis of the typical IPV relationship during adolescence. The methods which adult perpetrators use, such as socially isolating their victim and profusely apologizing after abuse, are well-known (Fall 2001). Potential victims can be educated and results of studies can be interpreted in light of the knowledge about the intimate relationship of a perpetrator and his victim. Such information, specific to adolescents and IPV, is not available for the average teenager. In-depth interviews and qualitative studies detailing the experiences of those involved in abusive relationships would be extremely helpful. Such knowledge would aid in the interpretation of findings from large studies such as the current study. In addition, vulnerable populations, such as adolescent mothers, could benefit substantially from more research in the dynamics of IPV. The current study has presented evidence that adult studies of IPV cannot be generalized to adolescents and that, instead, adolescent IPV needs to be uniquely understood.

Despite these limitations, the current study provides specific contributions to the field of IPV. By comparing IPV experiences between young adults and adolescents during the postpartum period, the present study has yielded information on an area which suffers from a dearth of research and a lack of understanding. The results are intriguing and against intuition, thereby strongly supporting the need for additional research in this area. The findings of this study contribute to knowledge regarding the unique experiences of abused adolescents, as well as the differences between young adult and adolescent mothers in the same situation.