Childhood Maltreatment, Posttraumatic Cognitions, and Health Outcomes Among Young Adults

Childhood maltreatment is a significant public health issue that has been linked to a myriad of negative long-term mental and physical health consequences. While the devastating health outcomes have been well established in the literature, the potential mechanisms of this link are less understood. In an effort to elucidate this relationship, the present study examined maladaptive cognitions (i.e., posttraumatic cognitions related to the self and world) that can form as a result of experiencing the trauma of maltreatment. Specifically, this cross-sectional study investigated the association between childhood maltreatment and self-reported mental and physical health concerns among a sample of young adults. Participants were 287 undergraduate students (ages 18-29 years) recruited from a mid-sized northeastern university. Retrospective, self-report questionnaires were used to assess childhood maltreatment, posttraumatic cognitions and current mental and physical health functioning. Data were analyzed using structural equation modeling. Childhood maltreatment was found to be significantly associated with poorer mental and physical health functioning. Support was found for posttraumatic cognitions as mediators in this relationship. Findings provide support for childhood maltreatment as an important risk factor for adverse long-term health outcomes, with posttraumatic cognitions playing an important role in this relationship. Maladaptive cognitions that form as a result of experiencing the trauma of maltreatment may be particularly useful points of intervention in order to mitigate health concerns in adulthood.


LIST OF TABLES
identified a wide range of negative sequelae, with particular attention to poorer long term mental and physical health outcomes (e.g., Briere & Elliot, 2003;WHO, 2006).
According to a recent report, the World Health Organization (WHO, 2006) declared childhood maltreatment to be associated with a range of health risk behaviors, directly contributing to some of the leading causes of death and chronic disease. Furthermore, while the actual physical injury associated with the traumatic event may be painful and cause harm to the child at the time of the event, the psychological consequences, longterm neurological, cognitive, and emotional development and overall health of the child are at risk for to be severely impacted as well (WHO, 2006). As a result, childhood maltreatment may be a significant etiological factor in the development and maintenance of health problems later in life.
The World Health Organization (WHO, 2006) distinguishes four types of childhood maltreatment that may occur before the age of 18: physical abuse, sexual abuse, emotional abuse, and neglect (i.e., physical and/or emotional). While childhood maltreatment is responsible for thousands of deaths each year, millions are also victims of non-fatal abuse and neglect (WHO, 2006). Those who are victims of childhood maltreatment, if untreated, are at high risk for developing poorer mental and physical health outcomes. As such, childhood maltreatment is a problem that touches communities worldwide, contributing to devastating health consequences.
Despite a clear link between childhood maltreatment and poorer overall health in adulthood, there is a surprisingly limited understanding of the intervening variables affecting this relationship. As such, an imperative next step is to examine these relationships to inform critical prevention and treatment efforts to reduce the prevalence and devastating effects of childhood maltreatment.

REVIEW OF LITERATURE
Childhood maltreatment is a large-scale problem that places a significant burden on the both the individual and the healthcare system. The effects of maltreatment can be immediate, but perhaps most importantly, can produce significant short-term and long-term sequelae (Buckingham & Daniolos, 2013). While the prevalence and total impact of childhood maltreatment is often difficult to determine due to underreporting and misdiagnosis, estimates based on confirmed cases of maltreatment point to an annual healthcare cost of 124 billion dollars in the United States alone (Buckingham & Daniolos, 2013;Fang, Brown, Florence, Mercy, 2012;Martin, Volkmar, & Lewis, 2007).
Prevalence rates of early maltreatment vary due to a number of complicating factors including, methodological issues (e.g., inconsistency in study design, definition of terms, study population, assessment measures), unreported or underreported cases, misdiagnosis (e.g., classified as another form of injury), and differing legal standards for substantiating maltreatment (Arnow, 2004;Buckingham & Daniolos, 2013;Macdonald, 2007;Rosenberg & Krugman, 1991;US Government Accountability Office, 2011). Despite these factors, documented prevalence rates are alarmingly high.
For example, in a community sample, researchers found that 37% of all respondents reported a history of either sexual or physical abuse before age 18 and approximately 21% of these respondents reported instances of both physical and sexual abuse (Briere & Elliot, 2003). In a primary care sample of adults, 44% reported a history of abuse during childhood (i.e., physical, sexual or emotional abuse) and 22% reported multiple forms of abuse (Gould et al., 1994). As may be expected, within a clinical sample, the highest rates of childhood maltreatment were reported (Briere, Woo, McRae, Foltz, Sitzman, 1997). According to the US Administration on Children and Families (2009), 3.3 million claims of maltreatment occur per year in the US and approximately 30% of these claims are substantiated. It is important to note that these statistics are likely an underestimation of actual prevalence rates, suggesting that even higher rates of maltreatment may exist. Consequently, the true prevalence of childhood maltreatment remains difficult to determine (Hussey, Change, & Kotch, 2006).
Keeping in mind the alarming rates of maltreatment, current findings suggest some differences in prevalence among men and women (e.g., Briere & Elliot, 2003;Finkelhor, 1994;Gould et al., 1994;Walker et al., 1999). For instance, females have been found to report more sexual abuse, emotional abuse, and multiple forms of abuse when compared to men (Miller, Green, Fettes, & Aarons, 2011). Consistent with these findings, a study examining undergraduates found that females reported more emotional abuse and sexual abuse than males (Paivio & Ramer, 2004). Evidence for gender differences in other forms of maltreatment, however, is mixed (Edwards, Holden, Felitti, & Anda, 2003;Miller et al., 2011;USDHHS, 2005).
The potential for multiple traumatic events during a child's life are far more likely to occur than a single instance of trauma (e.g., Felitti et al., 1998). Experiencing one form of childhood maltreatment increases the likelihood of experiencing other types of maltreatment, with approximately an 80 percent chance of experiencing another form of trauma in one's lifetime (Feltti et al., 1998). Studies considering the co-occurrence of physical, emotional and sexual abuse have found low prevalence rates (less than 10%) for single types of maltreatment, while the majority of victimized children are found to be victims of multiple forms of abuse (McGee, Wolfe, Yuen, Wilson, & Carnochan, 1995). Finally, evidence for racial and ethnic differences in the prevalence of child maltreatment has been found; however, findings are confounded by issues related to socioeconomic status and education level (Scher, Forde, McQuaid, & Stein, 2004).
The relationship between childhood maltreatment and poor mental health has been well established using clinical populations (e.g., Brown & Anderson, 1991;McCauley, Kern, Kolodner, Dill, & Schroeder, 1997) as well as community and population-based samples (e.g., Jumper, 1995;Neumann, Houskamp, Pollack, & Briere, 1996). Edwards and colleagues (2003) studied the effect of the co-occurrence of multiple forms of maltreatment (e.g., physical/sexual abuse, neglect) on psychological health among a sample of adult patients in an HMO (N=8,677).
Specifically, this study examined prevalence of various combinations of types of childhood maltreatment and explored whether the relationship between number of traumas and psychopathology was further impacted by emotional abuse in the childhood family environment. Results indicated that 34.6% of participants reported more than one type of trauma. A dose-response relationship between the number of indicators of childhood maltreatment and overall mental health scores was also supported, and the relationship between the number of traumas and poorer mental health was further strengthened by reports of an emotionally abusive family environment. Consistently, this type of graded relationship between number of traumatic experiences during childhood and poorer mental health outcomes in adulthood has been found (Mullen, Martin, Anderson, Romans, & Herbison, 1996;Felitti et al., 1998). In addition, gender differences are supported. Notably, males with a history of physical abuse were found to report more externalizing disorders in adulthood, whereas females tend to report more internalizing disorders, suggesting gender-specific consequences for the expression of psychiatric disorders (Keyes et al., 2012).

Adult Physical Health Outcomes
A growing body of literature suggests that childhood maltreatment can lead to poor physical health outcomes in adulthood. Adults who report a history of childhood maltreatment tend to have greater physical problems (both medically explained and unexplained) and tend to exhibit more high-risk health behaviors than individuals without a history of maltreatment (Walker et al., 1999;Felitti et al., 1998).
Perhaps most notably, such health concerns are some of the most common causes of death and disability in the United States (Felitti et al., 1998;White & Widom, 2003).
Survivors of childhood maltreatment have been found to be more likely to engage in risky health-related behaviors, and such behaviors may be precipitants or contributors to increased medical problems (Felitti et al., 1998;Walker, 1999;Chartier, Walker, & Naimark, 2009). In two separate studies, for example, women with a history of sexual abuse versus those without were found to report a greater number of health risk behaviors such as having multiple sexual partners, exchanging sex for drugs or money, using alcohol and drugs, and smoking (Coker, Smith, Bethea, King, & McKeown, 2000;Young & Katz, 1998). Similarly, additional studies have found that childhood abuse was significantly associated with several adverse physical health outcomes, including poorer perceived overall health, greater physical and emotional functional disability, increased numbers of distressing physical symptoms, and a greater number of health risk behaviors (e.g., Chartier, et al., 2009;Walker et al., 1999). This association was even stronger among females reporting a history of multiple forms of maltreatment. Felitti and colleagues (1998) suggest that an increase in health-risk behaviors might be related to using such behaviors (e.g., alcohol/drug use) as coping devices. Furthermore, a dose-response relationship between the number of instances of childhood maltreatment and the severity of negative adult physical health outcomes has been demonstrated, suggesting an additive effect of maltreatment (e.g., Arnow, Hart, Hayward, Dea, O'Connell, & Taylor, 1999;Felitti et al., 1998;Golding, Cooper, & George, 1997;Kessler, Davis, & Kendler, 1997;Sachs-Ericsson, Cromer, Hernandez, & Kendall-Tackett, 2009).

Health Care Utilization
As might be expected, childhood maltreatment has been identified as a key factor in predicting greater use of health care services (e.g., Chartier, Walker, & Naimark, 2010;Arnow et al., 1999;Yanos, Czaja, & Windom, 2010;Finestone et al., 2000;Newman et al., 2000;Walker et al., 1999). Studies of adult women with histories of child abuse, for instance, were found to report more annual primary care visits and have greater health care costs than individuals without histories of maltreatment (Arnow et al., 1999;Hulme, 2000;Walker et al., 1999). Similarly, significantly higher annual health care utilization and costs were found for women with history of child sexual and/or physical abuse; annual health care costs were 16% higher for women with sexual abuse only, 22% higher for women with physical abuse only and 36% higher for women with both types of abuse (Bonomi et al., 2008). Thus, childhood abuses are associated with long-term enhanced health care use and costs (Sarkar, 2010).
The relationship between maltreatment and higher rates of health care utilization may be best explained by this population's more significant psychopathology (Kessler, Sonnega, Bromet, Hughes, & Nelson, 1995) and increased medical comorbidity (Beckham et al., 1998;Elhai, North, & Frueh, 2005;Brown, Stout & Meuller, 1999). Population-based research has provided support for such findings (e.g., Chartier, Walker, & Naimark, 2009;Yanos, Czaja, & Widom, 2010). Similarly, young adults with a history of childhood maltreatment use significantly more health services compared to those with no history (Yanos, Czaja, & Widom, 2010). Arnow et al. (1999)  with a history of abuse were found to report significantly more doctors' visits than non-abused counterparts, with differences between groups on the objective measures following the same pattern. Interestingly, the study found only moderate correlations between self-report and the objective measures with a tendency to underreport service use (Newman et al., 2000).

Posttraumatic Cognitions
While rates of childhood maltreatment are disturbingly high and clearly associated with significantly poorer health outcomes, not all child victims go on to develop negative sequelae, raising the question of "why?" There is a limited understanding of the actual mechanisms that link childhood maltreatment to poorer overall health (e.g., Kendall-Tackett, 2003;Sachs-Ericsson, Cromer, Hernandez, Kendall-Tackett, 2009;Davis, Combs-Lane & Smith, 2004). In accordance with Foa and colleagues' emotional processing theory (Foa & Riggs, 1993;Foa, Steketee & Rothbaum, 1989), one pathway through which childhood maltreatment may lead to negative health is via the cognitions that form after experiencing maltreatment.
The emotional processing theory was developed to explain the development and maintenance of posttraumatic stress disorder (PTSD). This theory states that the relationship between trauma and PTSD is mediated by beliefs that the world is completely dangerous (belief about world) and that the self is completely incompetent (belief about self) and thus, PTSD is a result of a disruption in the "normal processes of recovery" (p. 303, Foa, Ehlers, Clarke, Tolin, & Orsillo, 1999). According to Foa and Rothbaum (1998), individuals may acquire these dysfunctional cognitions (i.e., "posttraumatic cognitions") one of two ways. On one hand, individuals may enter into the traumatic experience with rigid beliefs about the world being extremely safe and the self extremely competent, and thus, they have significant difficulty incorporating the traumatic experience into existing schemas. On the other hand, the event may act as a primer for pre-existing schemas of the world as dangerous place and the self as incompetent, especially among those who have experienced ongoing maltreatment. It is these individuals who are believed to be more vulnerable to developing PTSD. As such, individuals with less rigid concepts about the self and world may be more likely to acknowledge the maltreatment as a unique experience and less likely to develop PSTD.
Based on the emotional processing theory, the experience of maltreatment during childhood can impact victims' thoughts and beliefs by activating pre-existing negative schemas or by making it difficult to comprehend such a negative experience within a more rigid set of beliefs (e.g., Ehlers & Clarke, 2000;Foa & Riggs, 1993;Foa & Rothbaum, 1998;Foa, Ehlers, Clarke, Tolin, & Orsillo, 1999;Fraizer, 2003). As a result, the psychological effects related to childhood maltreatment may begin with cognitive distortions about the self and world that form after the event and then become incorporated into an individual's schemas at an early age (Smucker, Dancu, Foa, & Niederee, 2002). Since schemas are thought to develop from interactions within one's environment and then guide how individuals attend to and interpret their environment, a trauma during childhood could have a major impact on one's schema (Beck, 1967(Beck, , 1987Kovacs & Beck, 1978). Consequently, the role of the cognitive pathway, in which trauma-related cognitions may augment long-term symptomatology, has received increasing attention among researchers (e.g., Foa, Steketee, & Rothbaum, 1989;Foa & Riggs, 1993;McCann & Pearlman, 1990;Dunmore, Clark & Ehlers et al., 1999).
While the literature supports posttraumatic cognitions as a mediator in the relationship between traumatic experiences and PTSD, not all trauma victims go on to develop PTSD (e.g., Taft, Stern, King, & King, 1999;Wagner et al., 2000). In fact, one study found that while 90% of youth seen in a primary care clinic reported trauma exposure, only 25% met full or partial PTSD criteria (Lipschitz et al., 2000). A recent review of the research on children exposed to trauma found wide ranges in rates of PTSD among survivors of childhood maltreatment, noting studies found between 20 to 63% of survivors who met full criteria for PTSD (Gabbay, Oatis, Silva, & Hirsch, 2004). Therefore, it is possible that an individual's interpretation or appraisal of the traumatic event and its sequelae may play an important part in the development of subsequent psychopathology, not limited to PTSD (Foa et al., 1999;Elhers & Clarke, 2000). Posttraumatic cognitions may act to either buffer the negative effects of maltreatment or to prevent healthy coping (Foa & Rothbaum, 1998). While posttraumatic stress symptoms remain the most prevalent outcome for trauma survivors who experience cognitive distortions (Brewin & Holmes, 2003;Ehlers & Clark, 2000), dysfunctional cognitions are also highly associated with depressive symptoms (Coyne & Gotlib, 1983) and somatization symptoms (Rief, Hiller, & Margraf, 1998) in trauma survivors. Furthermore, such cognitions have been found to dramatically increase survivors' risk for disease and lead to poorer overall health (Kendall-Tackett & Klest, 2010).
Few extant studies have examined the long-term impact of posttraumatic cognitions among victims of childhood maltreatment (Gibb et al., 2001). Briere (2002) noted that one of the "earliest impacts of abuse and neglect is thought to be on the child's internal representations of self and other" (p.2). As a result, children who experience maltreatment tend to develop more negative views of self and others in the context of their maltreatment experience as compared to children who have not been victims of maltreatment (e.g., Allen & Tarnowski, 1989;Briere, 2002;Ponce, Williams, & Allen, 2004). In line with the emotional processing theory, these negative cognitions often include beliefs that the individual is "intrinsically unacceptable" or "deserving of such punishment or neglect" or may come to view him/herself as "helpless, inadequate, or weak" (Briere, 2002, p. 2). Furthermore, adults with a history of childhood maltreatment tend to report more negative cognitions as compared to those without a history of maltreatment (Gibb et al., 2001;Gold, 1986;Wenninger & Ehlers, 1998).
Some researchers have suggested that dysfunctional cognitions forming after childhood maltreatment may mediate the relationship between maltreatment and adult depression (Gibb et al., 2001). Using a college sample of men and women (N= 297), Gibb et al. (2001)  Furthermore, individuals' cognitive responses to early traumatic experiences and abilities to self-regulate emotions are thought to be important intervening variables though which early abuse contributes to physical health outcomes (e.g., Hagar & Runtz, 2012;Repetti, Taylor, & Seeman, 2002;Schnurr & Green, 2004).
While no studies to date have investigated the direct link between posttraumatic cognitions and poorer physical health, support for perceived stress and emotionfocused coping styles as partial mediators in the relationship between childhood maltreatment and adult physical health has been found (e.g., Hagar & Runtz, 2012).
Such findings suggest that survivors of childhood maltreatment tend to engage in emotion-focused coping strategies (e.g., efforts to manage internal experiences) (Hagar & Runtz). It is likely that these efforts at emotional coping are related to cognitions forming as a result of the maltreatment, as such coping strategies are typically in response to increased stress and emotional dysregulation post-trauma (Hagar & Runtz). In addition, thoughts related to poor self-esteem and selfcompetence have received some support in the literature as being related to poorer health, adding further support for the effect cognitions can have on one's health functioning (e.g., Benyamini, Leventhal & Leventhal, 2004;Ma'kikan-gas, Kinnuen, & Feldt, 2004).
Taken together, research studies indicate that the relationship between childhood maltreatment and poorer health-related outcomes is complex and suggests that maltreatment alone does not account for the variance in health outcomes. It is therefore important to consider mediating variables that may help to elucidate the well-established link between childhood maltreatment and the onset of mental and physical health problems. Cognitions forming after the experience of childhood maltreatment (i.e., "posttraumatic cognitions," "dysfunctional cognitions," or "trauma-related" cognitions), may play a role in future health outcomes. Thus, posttraumatic cognitions may be particularly relevant to understanding the relationship between childhood maltreatment and overall poorer mental and physical health in adulthood.
Clearly, experiencing maltreatment during childhood can have long-lasting effects on one's mental and physical health. As one might expect, the rates of health care utilization are also higher among this population when compared to those without a history of childhood maltreatment. Furthermore, the co-occurrence of multiple types of maltreatment makes it difficult to differentiate the effects of one type of maltreatment from another and a single occurrence of maltreatment is less likely than multiple instances (Finkelhor, Ormrod, Turner, & Hamby, 2005). Given the high rates of co-occurrence, studies focusing on only one form of maltreatment may overestimate its impact on long-term outcomes and obscure the effects of other forms of victimization (Hager & Runtz, 2012). Posttraumatic cognitions may be critical to understanding the effects maltreatment can have on poorer overall health and increased rates of health care utilization. However, to date, most of the research in this area is limited to the link between posttraumatic cognitions and the development and maintenance of PTSD.
In an effort to address gaps in the literature, the current study examined posttraumatic cognitions as a mediator in the relationship between childhood maltreatment and a broad set of health outcomes not limited to PTSD. In addition, this study examined this relationship in young adults with histories of childhood maltreatment whereas most of the literature has focused on older populations (i.e., middle to later adulthood). Specifically, this study examined whether posttraumatic cognitions mediate the relationship between childhood maltreatment and mental and physical health and rates of health care utilization (herein referred to as "Health Outcomes" for simplicity). Childhood maltreatment was examined as the independent variable and health outcomes (i.e., mental health, physical health, healthcare utilization) as dependent variables. Posttraumatic cognitions were evaluated as mediating variables.
Since the literature is suggestive of possible gender differences in type and frequency of maltreatment, as well as differences in mental and physical health outcomes, gender was initially considered as a moderator of this relationship.
Similarly, due to the due to the link between childhood maltreatment, posttraumatic cognitions, and the onset of PTSD, current PTSD symptomatology was also originally considered as a moderator. It would follow that the strength of the meditating effect of posttraumatic cognitions on childhood maltreatment-health outcomes link would depend on the presence or absence of current PTSD-symptomatology and gender.
However, due to limited sample size (see Results section for details), specific hypotheses regarding moderators could not be evaluated in the current study.

Hypotheses
This study aimed to test several models of relationships, including a mediation model, in a non-clinical sample of 287 college students. Specifically, the current study investigated the following hypotheses: 1. Childhood maltreatment is a predictor of poorer mental and physical health functioning.
2. Childhood maltreatment is a predictor of more negative posttraumatic cognitions.
3. More negative posttraumatic cognitions are related to poorer health outcomes (i.e., poorer mental/physical health functioning).
4. A mediational model will provide the best and most parsimonious fit to the data when comparing the following models (see Figures  is hypothesized that this model will be the most parsimonious model and provide the best fit to the data. 1c. Model C (Direct Model) specifies that the relationships between childhood maltreatment and mental and physical health outcomes are direct and not mediated by posttraumatic cognitions. It is hypothesized that the direct effects model will be least appropriate for both health outcomes.

Participants
Participants for the present study were drawn from a sample of 642 (158 males, 475 females and nine who did not report gender) college students. Since the current study intended to examine the relationships between childhood maltreatment, posttraumatic cognitions and health outcomes, only participants who endorsed a history of childhood trauma were included in the analyses described below. Of the 642 participants who completed the survey, 287 endorsed experiencing one or more types of childhood maltreatment (e.g., sexual, physical or emotional abuse, physical or emotional neglect) of at least "low to moderate" severity as determined by the Childhood Trauma Questionnaire (Bernstein & Fink, 1998). This sample consisted of 72 men (25.1%) and 215 women (74.9%) with an average age of 19.9 (SD = 2.4).
Participation was distributed among grade levels; 36.9% were freshman, 32.1% were sophomores, 18.5% were juniors, 10.8% were seniors and 1.7% did not report their year of education. The majority of participants were White (82.2%), while 4.9% were Black/African American, 4.5% were Asian, 11.8% identified as Hispanic/Latino, and 8.4% did not report their race/ethnicity. Most of the participants were unmarried (95.8%), over half (54%) lived on campus and nearly all had health insurance (91.3%).
About a third of the participants (34.5%) indicated a family history of mental health problems, whereas over half (53.3%) reported a family history of physical health problems. Participants were required to be at least 18 years old to participate in the study. See Table 1 for demographic characteristics.

Procedure
The study was approved by the Institutional Review Board (IRB) of the University of Rhode Island (IRB #HU1112-084). Participants were recruited from 100-level and 200-level courses (e.g., Introduction to Biology) during the Spring 2012 semester. Instructors were contacted and provided with information about the study.
Methods of recruitment varied based on instructors' preferences (e.g., attending class meetings to make an announcement, emailing enrolled students). Students who were at least 18 years old were eligible to participate.
Interested students were provided with a description of the study and a URL link to a secure and confidential online consent form via the online survey company, SurveyMonkey (www.surveymonkey.com). Participants were asked to read the consent form and click "yes" if they were willing to participate. If yes, participants were asked to complete a series of questionnaires. Due to SurveyMonkey's encryption software, participants' answers were never linked to any identifying information. The website remained accessible until the optimum sample size was reached. Participants were able to access the website from any computer at any time during data collection.
Participants were given the option to skip questions at any time. The survey took approximately 30-45 minutes to complete. At the end of data collection, the data was exported to the principle investigator's password-protected computer as an excel database.
Participants who completed the survey were eligible to be entered into a drawing for variety of gift cards (ranging from $10 to $50) by clicking on a link to a separate, secure online database and providing his or her email address. Study participation was voluntary and participants were given the investigator's contact information and encouraged to ask questions throughout the study. Participants were also allowed to withdraw from participation at any time.

Measures
Demographics. Participants were asked to complete a demographics questionnaire. Standard demographic information was requested from participants, including sex, age, race/ethnicity, and information about income, occupation, and education. Questions regarding participants' family demographic information were also asked including parents' highest level of education, family income, and family health history. Participants were also asked whether they currently had health insurance, and if yes, if it was the student health insurance package provided from URI.

Childhood Maltreatment.
Participants were asked to complete a childhood maltreatment measure, the Childhood Trauma Questionnaire (CTQ; Bernstein & Fink, 1998). The CTQ is a frequently used, reliable, and valid measure of maltreatment history (Bernstein et al., 2003). The CTQ is a retrospective, self-report instrument comprised of 28 items that assess five areas of childhood maltreatment using five subscales: Physical Abuse, Emotional Abuse, Sexual Abuse, Physical Neglect, and Emotional Neglect. Each item begins, "When I was growing up." Participants rate agreement with each item on a five-point Likert-type scale, where "1" represents never true for me and "5" represents very often true for me, according to the frequency with which the experiences occurred. With five items per subscale, scores range from 5-25 for each subscale.
The CTQ subscales, as well as the entire scale, have high test-retest reliability (r = .80-.88) and high levels of internal consistency (Cronbach's α = .79-.94). In the current study, the Cronbach alpha coefficient was .89. Validation studies support the CTQ's convergent and discriminant validity with structured interviews and corroboration with independent data. The CTQ has been used in studies of healthcare use and costs of primary care (Walker et al., 1999;Wright et al., 2001).
Subscale scores on each of the five subtests serve as indicators of childhood maltreatment. Cutoff scores for none to low, low to moderate, moderate to severe, and severe to extreme exposure are provided for each scale (Bernstein and Fink, 1998).
According to Bernstein and Fink (1998), these ranges were based on data from a nonclinical sample and were appropriate in identifying "cases" of specific types of abuse and neglect, while using therapist interview ratings for comparison. For the purposes of this study, the lowest level cutoff scores (i.e., low to moderate cutoff scores) for each subscale were used. Participants who endorsed at least low to moderate scores for one or more subscales of maltreatment were considered "positive" for a history of childhood maltreatment and retained for analyses. Cutoff scores were 9 or higher for emotional abuse, 8 or higher for physical abuse, 6 or higher for sexual abuse, 10 or higher for emotional neglect, and 8 or higher for physical neglect. A composite score was computed by summing subscale scores to serve as an indicator for childhood maltreatment.
Posttraumatic Cognitions. Participants completed a measure of posttraumatic cognitions, The Posttraumatic Cognitions Inventory (PTCI; Foa, Ehlers, Clark, Tolin, & Orsillo, 1999). The PTCI is a 36-item self-report measure with three subscales that assess for cognitions (i.e., beliefs about self and others) that can form as a result of experiencing a trauma. The measure is comprised of three subscales: negative cognitions about self (PTCI-Self, 21 items), negative cognitions about the world (PTCI-World, 7 items), and self-blame (PTCI-Blame, 5 items). For those participants who had reported a history of childhood maltreatment, participants were ask to rate the occurrence of thoughts and beliefs by indicating the degree to which they "agree" or "disagree" with each statement using a seven-point Likert scale ranging from 1 (totally disagree) to 7 (totally agree). The PTCI-Self measured the extent to which individuals had a negative view of him/herself and symptoms and thoughts of helplessness and alienation. The PTCI-World measured the degree to which individuals lack trust in others and believe the world to be unsafe. The PTCI-Blame scale measured the extent to which individuals attribute the occurrence of the event to something he/she did or did not do.
The three PTCI subscales and total score have demonstrated excellent internal consistency as follows: total score (α=.97); Negative Cognitions About Self, (α=.97); Negative Cognitions About the World (α=.88); Self Blame (α=.86) as well as good testretest reliability ranging from .74-.89 (Foa, Ehlers, Clark, Tolin, & Orsillo, 1999). In addition, the PTCI has demonstrated excellent convergent validity and discriminant validity (Foa, Ehlers, Clark, Tolin, & Orsillo, 1999 The PTCI total score was calculated by summing all items. Subscale scores are calculated as the mean item response for each subscale (Foa et al., 1999). In the current study, the PTCI subscales and total score demonstrated excellent internal consistency: McHorney, Ware, & Raczek, 1993). The SF-36 is a self-report questionnaire that is a reliable and valid measure of physical and mental health-related functioning. It measures health on eight multi-item dimensions, covering functional status, wellbeing and overall evaluation of health (Brazier et al., 1992). For each dimension, items scores are coded, summed, and transformed on a scale from 0 (worst health) to 100 (best health). The SF-36 is a generic measure, one that does not target a specific age group or disease, but instead, can be used with diverse populations. It was designed for use in surveys of general and specific populations, health policy evaluations, clinical practice and/or research.
The SF-36 has been widely used in clinical studies and has demonstrated adequate psychometric integrity (McHorney, Ware, & Raczek, 1993;Ware & Sherbourne, 1992). Factor analytic studies confirm that the eight scales assessing eight different health concepts make up two distinct factors: mental health and physical health, and that these factors account for 80-85% of the reliable variance in the eight scales in the U.S general population (Ware, Kosinski & Keller, 1994 visits; 3=7-9 visits; 4=10-12 visits; 5 = >12 visits) (Newman, 1998). Participants were asked: "Since September (six months ago), have you seen a health professional about your mental or physical health?" If yes, participants were asked to indicate the type of visit and frequency on the five-point Likert scale. This measure has been found to have high test-retest reliability over a 2-week period (Newman, 1998).

PTSD-Symptomatology.
Participants were asked to complete the Posttraumatic Diagnostic Scale (PDS), a measure of adult trauma and current PTSD symptomatology (Foa, Cashman, Jaycox, Perry, 1997). The PDS is a 49-item self-report measure that assesses severity of PTSD symptoms related to a single identified traumatic event. This symptomatology. Individuals were divided into two groups: 1) individuals who experienced a traumatic event with PTSD of at least moderate severity (score of 15 or above) and 2) individuals who experienced a traumatic event but did not have PTSD (PTSD-symptom severity score below 15) (Foa, Tolin, Ehlers, Clare, Orsillo, 1999

Preliminary Analyses
A Missing Values Analysis (MVA) was conducted to determine patterns of missing values (Tabachnick & Fidell, 2007). The PTCI scale had 6.6% missing data and the HCUQ scale had nearly 50% missing data; all other scales demonstrated less than 5% missingness. Due to the significant amount of missing data on the HCUQ (i.e., greater than 5%) and its nonrandom distribution, this measure was not used in the analyses (Tabachnick & Fidell, 2007). For the PTCI, however, the distribution of missing data was determined to be unpredictable and missing completely at random (MCAR), X 2 (46) = 52.63, p = .23. Since missing values that are distributed randomly throughout the data matrix pose less significant problems, those cases with missing data were not included in the main analyses (N=28) (Tabacknick & Fidell, 2007). All other questionnaires included in these analyses had less than 5% missing data.
Traditional maximum likelihood methods of structural equation modeling (SEM) assume that continuous variables in the model are normally distributed (Kline, 2010).
Preliminary analyses of univariate (e.g., means, standard deviations, skewness, and kurtosis) and multivariate normality indicated that the CTQ scores were positively skewed and highly kurtotic. For instance, 64.8% of the maltreated sample reported at least moderate levels of childhood maltreatment. Transformations were not made because it is reasonable to expect that these variables would be skewed in the population (i.e., most people do not experience moderate to severe instances of childhood maltreatment) (Tabachnick & Fidell, 2007). To take nonnormality into account, the decision was made to use ROBUST statistical methods, provided by EQS Structural Equation Modeling software. ROBUST methods assess chi square statistics and standard errors by use of maximum likelihood estimation with the Satorra-Bentler scaled chi square and adjust the standard errors to the extent of the nonnormality (Tabachnick & Fidell, 2007).
Mean, standard deviations and observed ranges for all variables are presented in Table 2. Based on the standards for categorization provided by Bernstein and Fink (1998), 44.7% of the original set of participants reported experiencing at least minimal levels of childhood maltreatment. See Table 3 for the percentage of participants endorsing each level of maltreatment for each type of maltreatment. All individuals included in the analyses endorsed at least one form of trauma meeting of at least minimal severity.

Exploratory Analyses
A series of independent samples t-tests were conducted on all dependent measures to test for gender differences. On most variables, gender differences were not found to be statistically significant. Females, however, were found to report significantly more emotional abuse on the CTQ than males ( Foa et al., 1999), whereas 117 did not endorse symptomatology. On most variables, significant group differences were found. See Table 4 for t-test results and effect sizes. While the magnitude of effect sizes ranged from small to large based on guidelines provided by Cohen (Tabachnick & Fidell, 2007), small sample sizes precluded measuring symptoms of PTSD as a moderator variable (Kline, 2010).  Table 5). Childhood maltreatment explained 13.4% of the variance in mental health outcomes, F (5,275) = 9.638, p < .001 (See Table 6), and 8.7% of the variance in physical health outcomes, F (5,275) = 6.342, p < .001 (See Table 7). The size and direction of the relationships indicate that childhood maltreatment is a significant predictor of poorer mental and physical health outcomes (Hypothesis 2). In particular, emotional, sexual and physical abuse contributed significantly to poorer mental and physical health outcomes. Tables 5-7 display the correlations between variables, unstandardized regression coefficients and standard errors. Based on these results, emotional and physical neglect were not supported as significant predictors of posttraumatic cognitions or mental and physical health outcomes.
The relationship between posttraumatic cognitions and mental and physical health outcomes was investigated using Pearson's correlation coefficient (Hypothesis 3). There was a significant negative correlation in the hypothesized direction between posttraumatic cognitions and mental health outcomes (r = -.59, p < 0.01). There was also a significant, negative correlation in the hypothesized direction between posttraumatic cognitions and physical health outcomes (r = -.19, p <0.01). Taken together, emotional, physical, and sexual abuse were supported as significant predictors of more negative posttraumatic cognitions and poorer health outcomes, and more negative posttraumatic cognitions were associated with poorer health outcomes.
Due to these findings, model modifications were made to remove emotional and physical neglect as indictors of the Childhood Maltreatment factor (see Figure 4). To meet MacKinnon's (2008) criteria for establishing mediation, methods assessing indirect effects in a mediational model were utilized. MacKinnon's method is preferable to the traditional causal steps method outlined by Baron and Kenny (1986) because the traditional test has been demonstrated to be underpowered (Fritz & MacKinnon, 2007). Structural equation model testing was conducted using the EQS 6.1 statistical software program (Bentler, 2003). Models were compared to one another in terms of their empirical fit with the data, using practical fit indices. Such indices included the comparative fit index (CFI; Bentler, 1990), the root mean square error of approximation (RMSEA; Steiger & Lind, 1998), R 2 values for estimating effect size, and standardized regression path coefficients for each model (Harlow, 2007). The CFI is a normed fit index that adjusts for degrees of freedom. CFI greater than 0.90, and RMSEA of less than .10 are indicative of good fitting models (Bentler & Wu, 1995). Insignificant chi-squares are ideal, however, due to the large sample size, it is expected that the chi-squares will be significant for the main analyses. If the macro fit indices for the model were deemed adequate, then the micro-fit indices were examined. These included the number of significant paths, their significance levels, and the strength of each relationship. For all analyses, a cutoff value for significance was set at p= 0.05 and ROBUST methods were used when evaluating fit.  In the current study, 44.7% of the initial sample (N=642) endorsed a history of childhood maltreatment of at least minimal severity. Among this maltreated group of participants, 62% percent reported a history of childhood emotional abuse, whereas 30% reported physical abuse and 25% reported sexual abuse. In addition, nearly 65%

Model A: Childhood Maltreatment Full
reported emotional neglect, while almost 39% reported physical neglect. Rates of maltreatment among male and female participants were relatively consistent with previous findings, with the exception of estimates of emotional and physical neglect, as these rates were slightly higher than some previous studies (e.g., Finkelhor, 1994;MacMillan et al., 1997;Bernstein & Fink, 1998). Importantly, however, rates of emotional, physical and sexual abuse were consistent with several previous studies which utilized identical CTQ subscales and criteria to estimate the prevalence of childhood maltreatment among undergraduate students (Bernstein & Fink, 1998;Turner & Paivio, 2002;Wright et al., 2001;see Paivia & Cramer, 2004 for details).
Also notable was the severity of maltreatment reported in the current study.

Participants were identified based on their responses on the Childhood Trauma
Questionnaire. Cutoff scores of "low to moderate" for each type of maltreatment were used to identify participants who had experienced one or more forms of childhood abuse and/or neglect (Bernstein & Fink, 1998) The lowest level cutoff scores were used in an effort to capture a high proportion of childhood maltreatment cases, including those of low severity (Paivio & Cramer, 2004 (Felitti et al., 1998;McGee, Wolfe, Yuen, Wilson, & Carnochan, 1995).
As hypothesized, childhood maltreatment was supported as a predictor of poorer mental and physical health outcomes among young adults. This finding is consistent with previous research demonstrating the strong association between childhood maltreatment and mental (e.g., Arnow, 2004;Banyard, Williams, & Siegel, 2001) and physical health difficulties (Walker at al., 1999;Felitti et al., 1998) in adulthood. More specifically, childhood abuse (i.e., sexual, physical and emotional abuse) accounted for 13.4% of the variance in mental health outcomes and 8.7% of the variance in physical health outcomes.
Childhood maltreatment was also found to be directly related to more negative posttraumatic cognitions. Similar to the health outcome findings, childhood abuse (i.e., emotional and sexual) accounted for the variance in cognitions. That is, young adults who reported experiencing childhood emotional and/or sexual abuse were more likely to report more negative cognitions associated with their early traumatic experiences. This is consistent with previous studies indicating that adults with a history of sexual and emotional abuse tend to report more negative cognitions as compared to those without histories of maltreatment (e.g., Wenninger & Ehlers, 1998).
Results supported the hypothesis that the relationship between childhood maltreatment and poorer health outcomes is mediated by posttraumatic cognitions.
Previous research has identified negative cognitions that form after experiencing childhood maltreatment as important variables in the relationship between child maltreatment and the onset of PTSD and depression (Gibb et al., 2001;Kayson, Scher, Mastnak, Resick, 2005). The current findings extend this line of research to more broadly include mental health and physical health functioning and support the mediating effect of trauma-related cognitions in this relationship. Since cognitions were not found to explain all of the variance in health outcomes, the remaining variance may be attributed to the direct effects of maltreatment (e.g., lingering effects of earlier injuries or infections) or to the impact of other variables that were not examined in this study (e.g., depression, PTSD, and other mental health conditions, revicitimization in adulthood, other childhood adversities including neglect, poverty, and poor nutrition, as well as health-risk behaviors and medical conditions, such as alcohol and tobacco use and obesity) (Hager & Runtz, 2012). It may also be due to measurement error and variability within the measures. The current study simultaneously examined the distinct forms of childhood maltreatment in an effort to control for the potential overlap among the different forms. Since the varying forms of maltreatment were not evaluated separately, the unique effects of emotional and physical neglect may have diminished when other forms of abuse were present, explaining the reason why abuse (e.g., sexual, emotional, and physical) emerged as the stronger predictor (Kayson, Scher, Mastnak, Resick, 2005). That being said, while the current study did not support physical and emotional neglect as strong predictors of health outcomes, it is possible they still contribute to the overall effect. Model fit for child abuse compared to models including neglect subscales were relatively similar. While it may be that the etiology of abuse is different than that of neglect, the relative stability in the two models suggests that neglect likely contributes similarly to overall outcomes. These findings point to the importance of studying multiple types of maltreatment simultaneously as it is not only difficult to parse out each form of maltreatment's effect, but also difficult to make sense of the interactions among the different forms leading to varying outcomes in adulthood (e.g., Edwards, Holden, Felitti, & Anda, 2003;Kayson, Scher, Mastnak, Resick, 2005).
The limited support for posttraumatic cognitions as a mediator between childhood maltreatment and physical health outcomes may be a result of several factors. First, college students are inherently a healthier population. In the current sample, participants rated their physical health more positively than their mental health. College students also tend to have higher socioeconomic status than the larger population, and thus, overall physical health is likely to be better. Age may also be a contributing factor (mean age = 19), as the types physical health-related issues assessed are less likely to emerge at such a young age. As such, it may be that physical health was less of a problem among this sample obscuring findings for physical health outcomes. Health risk behaviors (e.g., alcohol/drug use) may have also been more appropriate to assess among this population in an effort to target those behaviors that can lead to health problems later in life. It may also be that the types of cognitions assessed do not have as strong of a relationship with physical health and they do with mental health. It is possible that cognitions that form as a result of experiencing trauma may impact mental health functioning more strongly than physical health functioning. However, this seems less likely, as thoughts associated with self-esteem and competence have been linked to poorer health (e.g., Benyamini, Leventhal, & Leventhal, 2004;kikan-gas, Kinnunen, & Feldt, 2004) Furthermore, some gender differences were found. Women reported significantly more emotional abuse, poorer mental health functioning and more negative cognitions about the world as compared to men. However, the magnitude of differences was relatively small, which is possibly due to the small sample of men (N=72) versus women (N=215). As a result of the limited number of male participants, gender was not included as a moderator variable. The significant gender differences that were found, however, are consistent with previous research suggesting women's tendency to report more emotional abuse (Miller et al., 2011) and more negative cognitions than men (e.g., Ehlers & Clarke, 2003). Differences in mental health related issues may be in part due to women's tendency to report more psychological issues than men (Keyes et al., 2011). Taken together, it is possible that the above findings are best attributed to undergraduate men being less apt to participate in a study of this kind due to the nature of the questions (e.g., related to "previous childhood experiences"). Despite efforts to engage both men and women participants by reaching out to various majors on campus (e.g., Biology, Math, History) women were more likely to participate. Future studies should consider the efforts put forth in this study to reach both men and women, and include more targeted outreach approaches to increase the number of male participants.
Finally, this study examined the relationship between childhood maltreatment and health outcomes in a young adult, undergraduate student sample. It is believed that the consequences of childhood maltreatment would be particularly salient among young adults, as college marks a transition in the developmental process from childhood to adulthood. With this in mind, young adults who have experienced maltreatment as a child likely developed some degree of maladaptive cognitions in relation to the traumatic experience. The inherent stress that comes along with any major life transition (e.g., high school to college) may serve to exacerbate the effects of such negative cognitions and, thus, begin the development of mental and/or physical health related issues. Or, it may be that simply the age at which individuals typically enter college (18-20 years old) is the prime developmental period for the onset of such difficulties. More simply put, by examining a group of college students, this study was able to evaluate whether or not poorer mental and physical health outcomes were already emerging, at a relatively young age, among this group of maltreated individuals. Further, while the current study did not control for extraneous variables (e.g., repeat victimization as an adult), sampling such a young group of individuals reduces the likelihood that these variables have had an opportunity to impact participants' functioning. Thus, support for the mediating effects of posttraumatic cognitions among this group of college students suggests that as early as 18 years old, individuals with a history of childhood maltreatment are already experiencing negative health-related outcomes.

Limitations of the Present Study
Several methodological limitations of this study should be noted. First, the cross-sectional, correlational design of this study precludes conclusions about causality in the relationships between the variables. While most research in childhood maltreatment relies on retrospective reports, it is important to note the drawbacks of this method. A temporal sequence was artificially created in this study by asking participants to report traumatic experiences "prior to age 18" in addition to cognitions related to these experiences. However, while structural equation modeling (SEM) methods (compared to alternative methods) have the capability of inferring causality, definitive statements can only be made with longitudinal data (Lang, et al., 2008).
Since all variables in this study were assessed at a single time point, their potential roles as predictors, concomitants, or consequences cannot be determined with certainty. Therefore, neither casualty nor directionality can be assumed. Thus, it is impossible to determine whether negative trauma-related cognitions formed as a result of childhood maltreatment, resulted from, or caused more negative health outcomes.
Second, the present study assessed childhood maltreatment retrospectively and relied on participants' memories for information about childhood experiences. Such recollections may be subject to recall bias. Despite this limitation, research largely supports the accuracy of retrospective self-reports of childhood maltreatment (e.g., Brewin, Andrews, and Gotlib, 1993). Similar to the drawbacks of cross-sectional research (discussed above), retrospective self-report studies constitute the core of the literature on the long-term effects of childhood maltreatment. Longitudinal studies would allow for more accurate assessment of early childhood experiences and traumarelated factors.
An additional limitation is that the present study relied exclusively on selfreport measures. Therefore, the associations among variables could partly be accounted for by shared method variance, a threat to construct validity. Self-report tools are also vulnerable to effects of mood and social desirability (Hagar & Runtz, 2012). This may help to explain the limited data collected on health care utilization.
For example, it may have been that college students had difficulty remembering their last visit to the doctor (despite limiting the question to a 6-month period). It is also possible that participants had not been to the doctor in the past six-months, inherently limiting the number of participants who completed this questionnaire.
Generalizability is limited based on the homogenous nature of the current sample. Not only did the study focus exclusively on college students, the sample was limited in ethnic/racial and socioeconomic diversity. As such, it was composed of predominantly white, upper-middle class, undergraduate students which precluded testing for a number of potential demographic differences. Importantly, study findings may not generalize to other groups. It is possible that similar models may not find the same degree of fit with a more diverse population. Further, due to the tendency for college students to be a healthier subset of the overall population, it is likely that the current sample's mental and physical health functioning was higher than what would be found in a community sample. As such, future studies should aim to replicate findings in a more diverse group of young adults. In addition, studies should consider the impact of sociocultural factors on childhood maltreatment, posttraumatic cognitions, and long-term health outcomes. It would be particularly important to determine the role cultural and ethnic factors play in this relationship.

Implications of Study Findings
The results of the current study have significant implications for informing future studies and, perhaps most importantly, for improving prevention and treatment efforts. To begin, this study confirms the widespread occurrence of childhood maltreatment in a select sample of students who agreed to participate in a study about maltreatment and lends support for it as a major public health issue. In addition, study At minimum, it would seem to follow that more improved assessment measures for children who have experienced maltreatment are needed to not only target distress (e.g., depressive/anxious symptoms related to the trauma), but also, to target potential posttraumatic cognitions. Therefore, whether or not a child experiences distress early on, the child could at least receive interventions aimed at correcting dysfunctional cognitions, subsequently reducing their risk for later mental health issues. Such preventative efforts could also be aimed at increasing a child's cognitive flexibility, as more rigid concepts about the self and the world may place children at risk for developing subsequent psychopathology by making it difficult to incorporate the experience of a traumatic event into existing schemas (see Foa & Rothbaum's emotional processing theory, 1998).
There is also the possibility that unexamined variables account for some of the affects attributed to maltreatment, posttraumatic cognitions and health outcomes in the present model. Children who experience maltreatment also frequently experience other risk factors (e.g., low socioeconomic status, household dysfunction) which can further complicate a child's development (Appleyard, Egland, van Dulman, & Sroufe, 2005;Appleyard, Yang & Runyan, 2010). As a result, such factors may help to explain the development of posttraumatic cognitions. It is possible that these factors increase the risk of posttraumatic cognitions, as they may contribute to the formation of maladaptive schemas at an early age.

Future Directions
Based on study findings and limitations, future research should seek to clarify a number of questions raised by the current study. Due to the relatively homogenous sample, future studies should aim to evaluate this model in a more diverse group of young adults in an effort to enhance the generalizability of findings. In addition, the inclusion of both subjective (e.g., self-report, family/friends reports) and objective methods of data collection (e.g., records review), as well as, a lifetime clinical interview should be utilized to enhance the reliability of self-report measures.
Moreover, in an effort to understand the potentially devastating effects of childhood maltreatment, studies that are longitudinal in design are needed. Future prospective longitudinal research will allow for a more comprehensive understanding of the etiology of mental/physical health outcomes among survivors of childhood maltreatment and clarify the directionality of these relationships. Continued examination of these issues with multiple measures over time will help to clarify the developmental processes involved in the long-term effects of maltreatment. Finally, this study only begins to touch on the important mechanisms involved in the relationship between childhood maltreatment and health problems later in life.
Therefore, it is essential that future research continue to identify the mediating and moderating variables (e.g., stress level, family/social support, family dysfunction; SES) associated with this relationship in an effort to more effectively inform treatment and prevention efforts.
In summary, the present study supports the relationship between childhood maltreatment and health outcomes as mediated by posttraumatic cognitions among young adults. Thus, the manner in which young adults view themselves and the world around them as related to their early traumatic experiences appears to play a significant role in the later development of mental and physical health outcomes.
These findings support  Note. a N = 24 (8.4%) did not report race/ethnicity. b N= 5 (1.7) did not report education level. c N = 4 (1.4%) did not report marital status. d N= 2 (0.7) did not report parental level of education. e N = 22 (7.7%) did not report family income. f N = 2 (0.7) did not report mental health history.      You have been invited to take part in the research project described below. If you have any questions, please feel free to call or email Elizabeth Reichert, MA (978-979-0596; ereichert@my.uri.edu) or Ellen Flannery-Schroeder, PhD (401-874-4219; efschroeder@uri.edu), the people mainly responsible for this study.

Description of the project:
The purpose of this study is to better understand the long-term effects of childhood maltreatment. Child maltreatment includes any act that harms a person under the age of 18. Such acts of harm include: physical, emotional, or sexual abuse, and physical or emotional neglect. You may take part in this study whether or not you have ever experienced maltreatment of any kind.
Childhood maltreatment is often associated with many negative mental and physical health consequences. This study is attempting to better understand this link by learning more about what young adults think about themselves and others. We also want to learn about emotional difficulties young adults might be experiencing, like feeling nervous or sad, or physical health difficulties, such as experiencing pain or limited activity due to health problems.

What will be done:
If you agree to take part in this project, you will be one of about 700 college student who will be asked to complete the following survey that asks about your experiences as a child, perceptions about yourself and others, current emotional or physical health difficulties, and related subject matter.
To participate, you must be at least 18 years of age, be able to read and speak English, and access this study on a computer with internet access. This study is conducted entirely online and should take approximately 15-35 minutes to complete.
If you participate, you will have the option to be entered into a lottery for one of the following giftcards: five $10 iTunes giftcards, five $10 Dunkin' Donuts giftcards, and one $50 Visa giftcard. In order to be entered into the lottery, you will have to provide an email address but there will be no way to link your email address to your answers on the survey. At the end of the survey, you will click on a link to a separate survey, where you can submit your contact information for the lottery.

Possible risks or discomforts:
While the possible risks and/or discomforts of this study are minimal, it is possible for you to feel some distress or discomfort while answering questions about your experiences as a child, especially if you experienced maltreatment in the past.

Expected study benefits:
Although there are no direct personal benefits of the study, your answers will help increase the knowledge regarding the long-term effects of childhood maltreatment and what factors may play a role in the development of negative health consequences in adulthood. Importantly, this information will help to develop better interventions aimed at improving healthy functioning among young adults who have experienced maltreatment as a child.

Confidentiality:
Importantly, your part in this study is completely anonymous. That means that your answers to all questions are completely private. We will not know who provided which responses, and we will therefore not be able to trace your responses back to you.
No one else can know if you participated in this study and no one else can find out what your answers were. Scientific reports will be based on group data and will not identify you or any individual as being in this project. Your responses to assessment questions will be stored in a secure database on a server of the company that is hosting the internet survey (SurveyMonkey.com) and no identifying information will be included. We will not collect or store IP addresses. After online data collection is complete, the anonymous data will be transferred to the study investigator's computer and will remain password protected to ensure only the study investigator will have access to it.

Decision to quit at any time:
Taking part in this project is completely up to you. You do not have to participate and you can refuse to answer any question. You will be allowed to discontinue the survey at any time and you do not need to give any reasons for discontinuation. You are at least 18 years old. You have read the consent form and your questions have been answered to your satisfaction. You understand that you may ask any additional questions at any time and that your participation in this project is voluntary. Your filling out the survey implies your consent to participate in this study. If you would like a copy of this form, please print it out before continuing.

Rights and complaints:
Thank you in advance for your time, Elizabeth Reichert, MA

Your Health and Well-Being
This survey asks for your views about your health. This information will help keep track of how you feel and how well you are able to do your usual activities. Thank you for completing this survey! For each of the following questions, please mark an in the one box that best describes your answer.

1.
In general, would you say your health is: