Stress and traumatic stress: how do past events influence current traumatic stress among mothers experiencing homelessness?
Article Type: Survey
Subject: Post-traumatic stress disorder (Diagnosis)
Post-traumatic stress disorder (Social aspects)
Homeless women (Surveys)
Homeless women (Psychological aspects)
Mothers (Surveys)
Mothers (Psychological aspects)
Authors: Williams, Julie K.
Hall, James A.
Pub Date: 12/01/2009
Publication: Name: Social Work Research Publisher: National Association of Social Workers Audience: Academic; Trade Format: Magazine/Journal Subject: Sociology and social work Copyright: COPYRIGHT 2009 National Association of Social Workers ISSN: 1070-5309
Issue: Date: Dec, 2009 Source Volume: 33 Source Issue: 4
Topic: Event Code: 290 Public affairs
Geographic: Geographic Scope: Iowa Geographic Code: 1U4IA Iowa
Accession Number: 212686851
Full Text: The purpose of this research was to evaluate the relationship between past traumatic events and the level of current traumatic stress among mothers experiencing homelessness. The data for this study were gathered from 75 homeless mothers between May 2006 and October 2006 using a cross-sectional survey design with purposive sampling. All mothers were interviewed in a face-to-face, semistructured interview format using standardized questionnaires and measures, including the Global Appraisal of Individual Needs-Quick, Williams' Life History Calendar of Traumatic Events, the Traumatic Stress Index, and the Davidson Trauma Scale. The mothers ranged in age from 18 to 50. Forty-four percent were white, 21% were African American, 3% were Native American, 31% identified themselves as multiracial, and 9% reported Hispanic ethnicity. The analysis indicated that the average level of traumatic stress from past traumatic events and the number of distressing (but nontraumatic) events did not influence current traumatic stress; however, the number of past traumatic events significantly influenced the current level of traumatic stress among mothers experiencing homelessness. Recommendations for future research include investigating how traumatic stress affects a mother's ability to locate, find, and retain housing and how trauma interventions influence mothers to exit homelessness.

KEY WORDS: homeless mothers; Life History Calendar; posttraumatic stress disorder; stress; traumatic stress

**********

Between 30 percent and 40 percent of the estimated 3.5 to 4.5 million homeless people in the United States consist of families, mostly mothers with small children (Urban Institute, 2001; U.S. Conference of Mayors, 2007). Many of these mothers struggle daily to keep their children safe, warm, fed, and together as a family. At the least, these mothers have suffered the trauma of losing their home, a place of familiarity that provided a sense of security and grounding for them and their children. At the most, becoming homeless may have been the most recent in an accumulation of tragic events that these mothers have experienced. This study investigated traumatic stress among mothers experiencing homelessness, including whether past events influence current levels of traumatic stress among mothers.

Obtaining a better understanding of traumatic stress among mothers experiencing homelessness is important because of the debilitating effects traumatic stress can have on someone's ability to function. When a person experiences an intense traumatic reaction, the autonomic nervous system reacts in such a way that the higher functioning areas of the brain instinctively give way to the basic survival regions of the brain (Lanius et al., 2001). In effect, higher cognitive functioning is shut down or severely compromised. This automatic response can seriously impair decision-making capacities and communication skills of the traumatized individual. It is not surprising that this effect could have tremendous implications for homeless mothers who have experienced traumatic stress as they attempt to locate housing, hold down a job, or navigate social services on their own. Increasing our understanding of the prevalence of traumatic stress among homeless mothers and how past traumatic events may affect current traumatic stress could provide useful knowledge to social workers in their efforts to assist mothers who are homeless and suffering the effects of trauma. But what is traumatic stress? How is traumatic stress different from normal stress?

LITERATURE REVIEW

Traditional Stress Theory

Traditional stress theory originated with physician and pioneering researcher Hans Selye (Critelli & Ee, 1996; Pearlin, Lieberman, Menaghan, & Mullan, 1981). For Selye (1974), the main elements of stress theory included stressors and stress. Selye (1974) defined a stressor as anything that produces stress, with some stressors being pleasing (for example, the birth of a baby, a passionate embrace) and other stressors being unpleasant (for example, the loss of a job, having the flu); stress, however, is defined as "the nonspecific response of the body to any demand made upon it" (p. 27). Similar to stressors, stress can be pleasant or unpleasant. The term eustress refers to pleasant or beneficial stress, such as that which results from creative work or competition (Critelli & Ee, 1996); the term distress refers to unpleasant stress, such as that which results from receipt of bad news. According to Selye (1974), stress is both unavoidable and necessary in everyday living, as "complete freedom from stress is death" (p. 32).

The general idea of traditional stress theory is that living beings attempt to change and adapt to a continual barrage of stressors in an effort to reduce the effects of stress and restore a homeostatic state (Selye, 1974). During his research, Selye noticed that the body reacts to eustress and distress in the same way. First, the body experiences an alarm reaction. Next, the body enters a stage of resistance and attempts to cope with and adapt to the stress. Finally, if the stressor continues long enough, the body runs out of adaptive energy and becomes exhausted. Selye (1974) acknowledged that this cycle repeats itself throughout life "whenever we are faced with a demand" (p. 81). In addition, Selye noted that our reactions to stress are influenced by our individual adaptive energy stores, knowledge about the stressor, past experience, genetic and physical makeup, and environmental conditions.

While testing his theory in the laboratory, Selye discovered several important physiological processes that help the body restore homeostasis during stress (Vermetten & Bremner, 2002; Yehuda, 1998). His findings have been replicated countless times over the past half-century and have become so widely accepted that these physiological measurements are "considered de facto proof that stress ha[s] occurred" (Yehuda, 1998, p. 101). Consequently, when posttraumatic stress disorder (PTSD) gained recognition in the latter 20th century, researchers used their knowledge of stress and distress to formulate hypotheses for PTSD. Because previous research had shown that higher distress produced higher releases of certain chemicals, researchers expected to find a similar relationship between trauma severity and those chemicals (Yehuda, 2001). To their surprise, the results revealed the opposite, with individuals with PTSD having very different physiological reactions than individuals experiencing distress (Griffin, Resick, & Yehuda, 2005; Neylan et al., 2005;Yehuda et al., 2005; Yehuda, Gorier, Halligan, Meaney, & Bierer, 2004). Researchers have since found additional evidence that stress and traumatic stress have distinct etiologies (Yehuda, 1998).

Since these discoveries, behavioral patterns have been investigated to further differentiate stress from traumatic stress. According to Selye (1974), ordinary stress is unavoidable and necessary to everyday riving. If this is true, then behaviors related to stress could be deemed part of the usual behavior of people in everyday fife. When a person experiences traumatic stress, however, an unusual set of behaviors may occur. These behaviors may be identified as traumatic stress, which can lead to PTSD (American Psychiatric Association [APA], 2000).

PTSD

Traumatic stress encompasses the physiological, psychological, and behavioral reactions an individual has to an event that initially elicited feelings of intense fear, helplessness, or horror (APA, 2000). An event might be directly experienced (for example, violent physical assault, military combat), witnessed (for example, a severe car accident), or learned about (for example, learning that your child was sexually molested). If the physiological, psychological, and behavioral reactions last longer than a month and meet specific criteria, the individual could be suffering PTSD (APA, 2000). These criteria include exposure to an actual or perceived traumatic event in which the individual's response includes fear, helplessness, or horror; perceived re-experiencing of the event; avoidance of people, places, things, or situations associated with the event; hyperarousal; symptoms of re-experiencing, avoidance, and arousal lasting longer than one month; and substantial impairment in important areas of life functioning (APA, 2000). PTSD is considered acute if symptoms last less than three months and chronic if symptoms last at least three months, but it can also be delayed, with symptoms appearing at least six months after the event.

Few studies have investigated PTSD among homeless women, and only one study could be found that looked at, among other things, PTSD rates among homeless mothers. The Worcester Study (Bassuk, Buckner, Perloff, & Bassuk, 1998; Weinreb, Buckner, Williams, & Nicholson, 2006) investigated health, mental health, and substance abuse disorders of mothers who were homeless in Worcester, Massachusetts, in 1993 and 2003. In 1993, the past month rate of PTSD among the sample (n = 220) was 17.5%. By 2003, the rate among a sample of homeless mothers in Worcester (n = 148) had increased to 42.2%. These rates are substantially higher than the documented rate of lifetime PTSD among women in the general population, at 12% (APA, 2000). Additional studies are needed to corroborate the PTSD rates among mothers experiencing homelessness.

Multiple Traumas

Some researchers have suggested that multiple traumatic events may increase the level of traumatic stress among individuals (Brewin, Andrews, & Valentine, 2000; Resick, Yehuda, Pitman, & Foy, 1995). Only one peer-reviewed study could be located that specifically investigated the number of potentially traumatic lifetime events experienced by homeless mothers (Zugazaga, 2004), but the research considered only trauma exposure and did not include measures for traumatic stress. Nevertheless, the study found that the homeless women with children (n = 54) had endured an average of 12 exposures to potentially traumatizing life events, which is more than twice the number of exposures to potentially traumatic events in the general population, at five exposures (Breslau et al., 1998). Other research has suggested that a higher percentage of homeless women experience potentially traumatizing events compared with the general population. For example, studies have shown that homeless women experience significantly higher rates of adult victimization than do low-income-housed women (Kushel, Evans, Perry, Robertson, & Moss, 2003) and that homeless women are significantly more likely to have been sexually assaulted and have higher levels of distress than the low-income-housed women, (Ingram, Corning, & Schmidt, 1996). However, few studies have investigated multiple traumatic events among homeless mothers, and research could not be located that examined how multiple traumatic events influence current levels of traumatic stress among mothers experiencing homelessness.

The purpose of the present research was to evaluate the relationship between past traumatic events and the level of current traumatic stress among mothers experiencing homelessness. Specifically, this research examined the following three questions: (1) Does the number of past traumatic events affect the level of current traumatic stress among homeless mothers? (2) Does the average level of traumatic stress of past traumatic events affect the level of current traumatic stress among homeless mothers? (3) Does the number of past distressing (nontraumatic) events affect the level of current traumatic stress among homeless mothers?

METHOD

Participants and Setting

Participants for this study were recruited from mothers who had applied to Humility of Mary Housing, Inc. (HMHI), a nonsectarian social services agency that provides a supportive transitional and permanent housing program to homeless single-parent families. HMHI is located in Davenport, Iowa, the largest of four adjoining cities known collectively as the Quad Cities, and is the largest primary provider of supportive transitional and permanent housing to homeless single-parent families in the Quad Cities region. The agency was selected for data collection because HMHI receives 350 to 450 applications from homeless single-parent families per year. Participants for this study were eligible if they were female, were at least 18 years of age, were homeless, were the head-of-household for their family, were able to speak English, and had at least one child under the age of 18. All study procedures were approved by the Institutional Review Board at the University of Iowa.

Research Design

The data for this study were gathered between May 2006 and October 2006 using a cross-sectional survey design with purposive sampling. Seventy-five mothers were interviewed by a licensed, MSW-level researcher in a face-to-face, semistructured interview format using standardized questionnaires and measures. This format allowed the interviewer to watch for and address any negative reactions to the questionnaires or measures, such as excessive psychological discomfort or trauma-related symptoms (for example, dissociation, re-experiencing).Throughout the interview, the interviewer used trauma-informed practices, such as preparing the mother for potentially difficult questions and reminding the mother that she was in control of the interview, could skip any questions or assessments, or stop the interview at any time. In all cases, mothers completed the entire interview.

Assessment Procedures

After the study was explained and the consent document signed, the Global Appraisal of Individual Needs--Quick (GAIN-Q) (Dennis, Titus, White, Unsicker, & Hodgkins, 2002) was introduced and completed, with the interviewer reading the questions and marking the responses. Williams' Life History Calendar (LHC) (Williams, 2007) was then introduced and completed. The LHC approach was chosen over a regular interview or survey approach because research has shown that the LHC helps respondents to provide more detailed and accurate retrospective data than either a structured interview or a survey method (Freedman, Thornton, Camburn, Alwin, & Young-DeMarco, 1988; Lin, Ensel, & Lai, 1997; Yoshihama, Gillespie, Hammock, Belli, & Tolman, 2005).

Next, the mother was asked to identify events in her life that were particularly upsetting to her. The interviewer listed events named or described by the mother in a separate section on the LHC. The mother was then asked to identify the event from the list that was currently most disturbing to her. This event was used to measure the mother's current level of traumatic stress using the Davidson Trauma Scale (DTS) (Davidson, 1996). After the DTS was completed, the participant completed a Traumatic Stress Index (TSI) (Dennis, 1998) for each of the events listed. At the end of the interview, a collection of counseling referrals was reviewed and the mother was compensated $20 for her time and thanked for participating in the study.

Measures

GAIN-Q. The GAIN-Q (Dennis, Titus, White, et al., 2002) is an abbreviated version of the comprehensive GAIN (Dennis, 1998), which is a reliable and valid behavioral health screening instrument (Buchan, Dennis, Tims, & Diamond, 2002; Dennis, Funk, Godley, Godley, & Waldron, 2004; Dennis, Titus, Diamond, et al., 2002) with norms available for adults (Dennis, Scott, Godley, & Funk, 1999). The GAIN-Q is organized into nine main sections relating to demographic information (for example, race and ethnicity, gender, age, years of education), past-year physical and mental health, past--year substance use, and service utilization. The GAIN-Q can be administered in approximately 15 minutes.

Williams' LHC. Williams' LHC was created by the first author specifically for the broader study and follows the general format of other LHCs (Freedman et al., 1988; Lin et al., 1997; Yoshihama et al., 2005). The calendar is in the form of a large grid in which the columns represent time (participant lifetime in years) and the rows represent different domains, activities, or events of interest. Domains included residence, education, children, employment, substance use, physical health, mental health, and abuse history. For each domain, the interviewer asked the respondent when an event occurred and for how long, and the interviewer filled in the appropriate cells of the grid. Codes were used to delineate specific information (for example, if the row represented education, the letter Q or G in a cell would designate when the respondent quit or graduated, respectively). This process continued until all of the appropriate rows of the grid had been addressed.

DTS. The DTS (Davidson, 1996) is a 17-item assessment instrument for adults that measures the frequency and severity of past-week PTSD symptoms according to Diagnostic and Statistical Manual of Mental Disorders (4th ed., text rev.) (DSM-IV-TR) (APA, 2000) criteria for one identified event. Specifically, items 1 to 4 and item 17 match criteria B for PTSD (recurrent images, thoughts, nightmares related to the event), items 5 to 11 match criteria C (avoidance and emotional numbing), and items 12 to 16 match criteria D (hyperarousal). Items are measured on a scale ranging from 0 to 4 for both frequency (not at all, once only, 2 to 3 times, 4 to 6 times, every day) and severity (not at all distressing, minimally distressing, moderately distressing, markedly distressing, extremely distressing). Example questions on the DTS include "Have you had distressing dreams about the event?"" Have you been upset by something that reminded you of the event?" and "Have you been avoiding thoughts or feelings about the event?" The highest possible score on the DTS is 136, with 40 as the cut point score for clinical PTSD (Davidson et al., 1997). The DTS takes less than 10 minutes to complete.

The DTS has been tested in both general and clinical populations and across ethnic groups (Davidson et al., 1997; Davidson, Tharwani, & Connor, 2002). The DTS has high internal consistency (Cronbach's alpha ratings are .97 and .98 for the frequency and severity scales, respectively, and .99 for all 17 severity and frequency items together) (Davidson et al., 1997).The DTS has shown good convergent and discriminant validity and has proven to be significantly correlated (p < .001) with the Clinician-Administered PTSD Scale (.78) (Blake et al., 1995), the Impact of Event Scale (.64) (Weiss, 2004), and the Symptoms Checklist 90-Revised (.89) (Davidson et al., 1997). The DTS was chosen over other measures because it is brief, has good reliability and validity, and has been tested in both general and clinical populations and across ethnic groups.

TSI. The TSI is an internally consistent (Cronbach's alpha = .92) 13-item scale that counts the number of traumatic stress symptoms related to memories of past traumatic events. The scale is based on a subset of items from the Civilian Mississippi Scale for PTSD (Norris & Perilla, 1996; Vreven, Gudanowski, King, & King, 1995), excluding the reverse-scored items deemed problematic in prior research (Scott, Sonis, Creamer, & Dennis, 2006). TSI items relate to the DSM-IV-TR (APA, 2000) criteria for PTSD, including symptoms of re-experiencing (for example, "You had nightmares about things in your past that really happened"), emotional numbing (for example, "It seemed as if you had no feelings"), increased arousal (for example, "You lost your cool and exploded over minor, everyday things"), and impairment in functioning (for example, "You felt like you could not go on"). Previous research indicates concurrent validity between the TSI and blind psychiatric diagnoses of traumatic disorders (Jasiukaitis & Shane, 2001). Because scores of 5 or higher indicate traumatic stress levels associated with a traumatic stress disorder (Chestnut Health Systems, 2002), events with scores of 5 or higher were categorized as traumatic events. Events with scores lower than 5 were categorized as distressing events. Norms and confirmatory factor analysis are available (Chestnut Health Systems, 2002).

Data Analysis Procedures

The data were analyzed using STATA SE 8 (Statacorp, 2003). The research questions were investigated using ordinary least-squares multiple regression techniques. For the multiple regression model, the dependent variable was the DTS score, which measured the current level of traumatic stress. The independent variables in the model included number of past traumatic events, average TSI score of past traumatic events, and number of past distressing events. The participant's age at the time of the interview was controlled because participant age could influence the number of events experienced, and the number of months since the DTS event was controlled because some mothers identified recent events and others identified events that occurred decades ago. All regression diagnostic testing, including tests examining multicollinearity (mean variance inflation factor = 1.52, minimum [min] = 1.16, maximum [max] = 2.05), model specification using Ramsey's regression specification error test IF(3, 66) = 0.76,p = .52] and the link test (hatsq = 0.005, p = .28), nonlinearity (using scatter plots of the residual against predictor variables; contact first author for plots), heteroscedasticity (Bruesch-Pagan/ Cook-Weisberg [chi square] = 1.77, p = .18), normality of residuals (Shapiro-Wilk z = 0.17, p = .43), and influential outliers (leverage versus squared residual plot; contact first author for plot) indicated that the assumptions of the model were met.

RESULTS

The sample consisted of 75 homeless mothers ranging in age from 18 to 50 (M = 30 years, SD = 8.1 years). Forty-four percent of the mothers were white, 21% were African American, 3% were Native American, 31% identified themselves as multiracial, and approximately 9% reported Hispanic ethnicity. Eighty percent of the mothers had at least a high school diploma or equivalent. Seventy-six percent of the mothers had either worked or attended school in the past 90 days, with those mothers working an average of 45 of the past 90 days. Fifty-five percent of the mothers had never been married, 16% were married but separated, and 29% were divorced. Eighty-nine percent of the mothers were living with their children at the time of the interview, with the remaining mothers reporting plans to reclaim physical custody of their children after locating stable housing. The number of lifetime homeless episodes ranged from one to 10, with a median of two episodes.

Current Level of Traumatic Stress

Mothers identified a single event that was most disturbing at the time of the interview, and this event was used to measure their current level of traumatic stress. Sixty-one percent of mothers identified an event that occurred within the past year, and 71% identified an event that occurred within the past two years. The most often cited event involved abuse (for example, physical, sexual, emotional). Nearly 67% of the mothers reported scores of 40 or higher on the DTS (M = 56.2, Mdn = 56.0, SD = 33.5, min = 0, max = 125), thus indicating that nearly two-thirds of the mothers were suffering PTSD at the time of the interview.

Fifty-two percent of the mothers identified an event that occurred prior to becoming homeless, with 72% of those having a DTS score indicating PTSD. Of the 44% identifying an event that occurred during the current homeless episode, 61% had a DTS score indicating PTSD. No differences were found between mothers identifying their most disturbing event before or after becoming homeless among DTS scores (t = -0.28,p = .79) or presence of PTSD [[chi square] (1, N = 75) = 0.96,p = .33].

Past Events

Mothers identified an average of 12 total events (SD = 6.25, min = 0, max = 26). Of those, a median of eight events were traumatic (that is, events with TSI scores of 5 or higher; M = 8.5, SD = 5.7), with a median average traumatic score of 7.8. The average number of distressing events (that is, events with TSI scores below 5) was 3.25 (SD = 2.51, min = 0, max = 16).

Association of Past Events on Levels and Current Traumatic Stress

The research questions were investigated using a multiple regression model that regressed the mother's current traumatic stress score on the number of past traumatic events, average traumatic score on past traumatic events, and the number of past distressing events, controlling for age and number of months since the DTS event (that is, the event used to measure the current level of traumatic stress) occurred. The overall model was significant [F(5, 69) = 10.18, p < .001] and explained 38% of the variance in the mothers' level of current traumatic stress (see Table 1).

The analysis indicated that as the number of past traumatic events increased, the levels of current traumatic stress among the mothers also increased, holding all else constant; specifically, for every additional past traumatic event the mother experienced, the level of current traumatic stress increased by nearly three points (p < .001), holding the average level of traumatic stress of past traumatic events and the number of distressing events constant. In fact, for every standard deviation increase in the number of past traumatic events (SD = 5.7 events), the level of current traumatic stress increased by more than 15 points on the DTS, holding the average score of past traumatic events and the number of distressing events constant. For the remaining research questions, the average level of traumatic stress of past traumatic events and the number of distressing (but nontraumatic) events was not associated with the level of current traumatic stress among the mothers (see Table 1).

DISCUSSION

The high number of homeless families in the United States suggests that more work is needed in understanding the problems and difficulties that homeless families face. One potential problem that has received little attention in the research literature is traumatic stress among mothers experiencing homelessness. Because high traumatic stress is known to have debilitating effects on the lives of people, and because prior research indicates that many homeless mothers have experienced difficult and potentially traumatic events prior to becoming homeless, past events may influence the level of current traumatic stress among homeless mothers. The purpose of this research was to explore the relationship between past traumatic events and the level of current traumatic stress among mothers experiencing homelessness.

This study found that more than two-thirds of the mothers in the sample were suffering from PTSD at the time of the interview and that most mothers had suffered multiple traumatic events during their lifetimes. When asked to identify the event that was most disturbing to them at the time of the interview, the majority of mothers identified an event that occurred before they became homeless. That the majority of mothers perceived events that occurred prior to homelessness as more disturbing than either homelessness or events that occurred during homelessness underscores the impact of past traumatic events on mothers experiencing homelessness. In addition, mothers in the study had experienced an average of 12 distressing or traumatic events in their lifetime--which is more than twice the number of events identified by women in general community samples (Breslau et al., 1998)--and more than two-thirds of those events could be categorized as traumatic events. Thus, a contribution of this research is the effort to press beyond prior research (which considered only trauma exposure) by attempting to measure whether the past events were distressing or traumatic for the mother.

The results of differentiating between traumatic and distressing events were revealing, as the number of past traumatic events significantly increased current traumatic stress, but the number of past distressing (that is, nontraumatic) events had no influence. Hence, differentiating between traumatic and distressing past events was important to assessing the potential influence of past events, with significance resting specifically with the number of past traumatic events. This is important because it demonstrates that trauma exposure does not automatically equate to traumatic stress. Therefore, future research should differentiate between distressing and traumatic past events.

Implications for Social Work Practice

Overall, the current findings support existing theory and research in addition to advancing our knowledge of traumatic stress among mothers experiencing homelessness. Such findings are important to advance future research and clinical practice. Specific to clinical practice, social workers assisting homeless mothers must attend to the needs relating to homelessness; however, the fundamental principle of social work practice of beginning where the client is suggests that workers should also help mothers cope with the traumas they have experienced. Social workers working with people experiencing homelessness should assume that all people experiencing homelessness have been exposed to trauma; therefore, social workers should receive training in and make use of trauma-informed and trauma-sensitive practice.

These findings also have implications for social work administrators. Administrators of social service agencies providing services to the homeless population are often forced to focus primarily on housing and employment issues, largely due to funding limitations and related political ideologies. Because high levels of traumatic stress may interfere with a mother's ability to maintain stable housing or employment and because prior research indicates that the earlier traumatic stress is addressed the better the outcomes for traumatized individuals (Langill, Ingargiola, Schwartz, & Kutyla, 2005), this study provides evidence that may help administrators justify funding requests and budgetary changes for staff training in early trauma identification and for developing appropriate services to address traumatic stress. These changes could substantially improve the chances for long-term housing and employment of homeless mothers.

Social work researchers and practitioners must work together to develop, implement, and evaluate evidence-based trauma interventions appropriate for mothers experiencing homelessness. Although much work needs to be done, social workers can help mothers who are traumatized and experiencing homelessness through rigorous research and appropriate social work practice skills. Together, social work researchers, practitioners, and their allies can advance the knowledge base to develop ways to help traumatized mothers experiencing homelessness.

Strengths and Limitations of the Research

The strengths of this research include the face-to-face semistructured interview format, which allowed the interviewer to watch for and address any negative reactions to the questionnaires or measures, such as heightened psychological discomfort or trauma-related symptoms (for example, dissociation, re-experiencing), and to reduce the incidence of missing or incomplete response sets. This research also has limitations. Because homeless mothers are generally a hidden population, generalizing the responses from this sample to the larger population of homeless mothers is difficult. The representativeness of the sampling frame is unknown for the nationwide female homeless population. Also, purposive sampling limits the generalization of the study, as does a single site study with a smaller sample.

CONCLUSION

This study investigated traumatic stress among mothers experiencing homelessness, including how past events were associated with increased current levels of traumatic stress among the mothers. Although the average level of traumatic stress of past traumatic events and the number of distressing (but nontraumatic) events did not influence current traumatic stress among the mothers, the number of past traumatic events was associated with the current level of traumatic stress among mothers experiencing homelessness. These findings indicate that differentiating between distressing and traumatic past events may improve the precision and validity of the research. Recommendations for future research include measuring the level of traumatic stress for past events to strengthen the rigor of research with traumatized mothers experiencing homelessness, investigating how traumatic stress affects a mother's ability to locate, find, and retain housing and how trauma interventions influence the progress of mothers to exit homelessness.

Original manuscript received January 22, 2008

Final revision received November 17, 2008

Accepted January 20, 2009

REFERENCES

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author.

Bassuk, E. L., Buckner, J. C., Perloff, J. N., & Bassuk, S. S. (1998). Prevalence of mental health and substance use disorders among homeless and low-income housed mothers. American Journal of Psychiatry, 155, 1561-1564.

Blake, D. B., Weathers, F.W., Nagy, L. M., Kaloupek, D. G., Gusman, F. D., Charney, D. S., & Keane, T. M. (1995). The development of a Clinician-Administered PTSD Scale. Journal of Traumatic Stress, 8, 75-90.

Breslau, N., Kessler, R. C., Chilcoat, H. D., Schultz, L. R., Davis, G. C., & Andreski, P. (1998).Trauma and posttraumatic stress disorder in the community. Archives of General Psychiatry, 55, 626-632.

Brewin, C. R., Andrews, B., & Valentine, J. D. (2000). Meta-analysis of risk factors for posttraumatic stress disorder in trauma-exposed adults. Journal of Consulting and Clinical Psychology, 68, 748-766.

Buchan, B.J., Dennis, M. L., Tims, F. M., & Diamond, G. S. (2002). Cannabis use: Consistency and validity of self-report, on-site urine testing and laboratory testing. Addiction, 97(Suppl. 1), 98-108.

Chestnut Health Systems. (2002). GAIN-I scales and variables. Retrieved June 11, 2006, from http://www. chestnut.org/LI/gain

Critelli, J.W., & Ee, J. S. (1996). Stress and physical illness: Development of an integrative model. In T.W. Miller (Ed.), Theory and assessment of stressful life events (pp. 139-159). Madison, CT: International Universities Press.

Davidson, J.R.T. (1996). Davidson Trauma Scale. North Tonawanda, New York: Multi-Health Systems.

Davidson, J.R.T., Book, S.W., Colket, J. T., Tupler, L.A., Roth, S., David, D., et al. (1997). Assessment of a new self-rating scale for post-traumatic stress disorder. Psychological Medicine, 27(1), 153-160.

Davidson, J.R.T., Tharwani, H. M., & Connor, K. M. (2002). Davidson Trauma Scale (DTS): Normative scores in the general population and effect sizes in placebo-controlled SSRI trials. Depression and Anxiety, 15, 75-78.

Dennis, M. L. (1998). Global Appraisal of Individual Needs (GAIN). Bloomington, IL: Chestnut Health Systems.

Dennis, M. L., Funk, R., Godley, S. H., Godley, M. D., & Waldron, H. (2004). Cross-validation of the alcohol and cannabis use measures in the Global Appraisal of Individual Needs (GAIN) and Timeline Followback (TLBF, Form 80) among adolescents in substance abuse treatment. Addiction, 99(Suppl. 2), 120-128.

Dennis, M. L., Scott, C. K., Godley, M. D., & Funk, R. (1999). Comparison of adolescents and adults by A SAM profile using GAIN data from the Drug Outcome Monitoring Study (DOMS): Preliminary data tables. Retrieved May 30, 2006, from http://chestnut.org/ LI/Posters/asamprof.pdf

Dennis, M. L., Titus, J. C., Diamond, G., Donaldson, J., Godley, S. H., & Tiros, F. M. (2002). The Cannabis Youth Treatment (CYT) experiment: Rationale, study design and analysis plans. Addiction, 97(Suppl. 1), 16-34.

Dennis, M. L., Titus, J. C.,White, M. K., Unsicker, J. I., & Hodgkins, D. (2002). Global Appraisal of Individual Needs (GAIN): Administration guide for the GAIN and related measures. Bloomington, IL: Chestnut Health Systems.

Freedman, D., Thornton, A., Camburn, D., Alwin, D., & Young-DeMarco, L. (1988). The Life History Calendar: A technique for collecting retrospective data. Sociological Methodology, 18, 37-68.

Griffin, M. G., Resick, P.A., & Yehuda, R. (2005). Enhanced cortisol suppression following dexamethasone administration in domestic violence survivors. American Journal of Psychiatry, 162, 1192-1199.

Ingram, K. M., Corning, A. F., & Schmidt, L. D. (1996). The relationship of victimization experiences to psychological well-being among homeless women and low-income housed women. Journal of Counseling Psychology, 43, 218-277.

Jasiukaitis, R, & Shane, P. (2001). Discriminant analysis with GAIN-I psychological indices reproduces staff psychiatric diagnoses in an adolescent substance abusing sample. Paper presented at the Persistent Effects of Treatment Study of Adolescents Analytic Cross-site Meeting, Washington, DC.

Kushel, M. B., Evans, J. L., Perry, S., Robertson, M.J., & Moss, A. R. (2003). No door to lock: Victimization among homeless and marginally housed persons. Archives of Internal Medicine, 163, 2492-2499.

Langill, D., Ingargiola, R, Schwartz, A., & Kutyla, T. (2005). In harm's way: Aiding children exposed to trauma. Denver: Grantmakers in Health.

Lanius, R. A., Williamson, R C., Boksman, K., Gupta, M.A., Neufeld, R.W., Gati, J. S., et al. (2001). Neural correlates of traumatic memories in posttraumatic stress disorder: A functional MRI investigation. American Journal of Psychiatry, 158, 1920-1922.

Lin, N., Ensel, W. M., & Lai, W. G. (1997). Construction and use of the life history calendar: Reliability and validity of recall data. In I. H. Gotlib & B. Wheaton (Eds.), Stress and adversity over the life course (pp. 249-272). New York: Cambridge University Press.

Neylan, T. C., Brunet, A., Pole, N., Best, S. R., Metzler, T.J., Yehuda, R., et al. (2005). PTSD symptoms predict waking salivary cortisol levels in police officers. Psychoneuroendocrinology, 30, 373-381.

Norris, F. H., & Perilla, J. L. (1996).The revised Civilian Mississippi Scale for PTSD: Reliability, validity, and cross-language stability. Journal of Traumatic Stress, 9, 285-298.

Pearlin, L. I., Lieberman, M.A., Menaghan, E. G., & Mullan, J.T. (1981). The stress process. Journal of Health and Social Behavior, 22, 337-356.

Resick, H. S., Yehuda, R., Pitman, R. K., & Foy, D.W (1995). Effects of previous trauma on acute plasma cortisol level following rape. American Journal of Psychiatry, 152, 1675-1677.

Scott, C. K., Sonis, J., Creamer, M., & Dennis, M. L. (2006). Maximizing follow-up in longitudinal studies of traumatized populations. Journal of Traumatic Stress, 19, 757-769.

Selye, H. (1974). Stress without distress. Philadelphia: J. B. Lippincott.

Statacorp. (2003). Stata/SE 8.0 [Computer software]. College Station, TX: Author.

Urban Institute. (2001). A new look at homelessness in America. Retrieved November 16, 2003, from http:// www.urban.org/urlprint.cfm?ID=6814

U.S. Conference of Mayors. (2007). Hunger and homelessness survey: A status report of hunger and homelessness in America's cities. Washington, DC: Sodexho, Inc.

Vermetten, E., & Bremner, D. (2002). Circuits and systems in stress. Depression and Anxiety, 15, 126-147.

Vreven, D. L., Gudanowski, D. M., King, L. A., & King, D.W (1995).The civilian version of the Mississippi PTSD scale: A psychometric evaluation. Journal of Traumatic Stress, 8, 91-109.

Weinreb, L. F., Buckner, J. C., Williams, V., & Nicholson, J. (2006). A comparison of the health and mental health status of homeless mothers in Worcester, Mass: 1993 and 2003. American Journal of Public Health, 96, 1444-1448.

Weiss, D. S. (2004).The Impact of Event Scale-Revised. In J. P. Wilson & T. M. Keane (Eds.), Assessing psychological trauma and PTSD (pp. 399-411). New York: Guilford Press.

Williams, J. K. (2007). Williams' Life History Calendar. Iowa City: University of Iowa.

Yehuda, R. (1998). Neuroendocrinology of trauma and posttraumatic stress disorder. In R. Yehuda (Ed.), Psychological trauma (pp. 97-131).Washington, DC: American Psychiatric Press.

Yehuda, R. (2001). Biology of posttraumatic stress disorder. Journal of Clinical Psychiatry, 62(Suppl. 17), 41-46.

Yehuda, R., Engel, S. M., Brand, S. P,., Seckl, J., Marcus, S. M., & Berkowitz, G. S. (2005). Transgenerational effects of posttraumatic stress disorder in babies of mothers exposed to the World Trade Center attacks during pregnancy. Journal of Clinical Endocrinology & Metabolism, 90, 4115-4118.

Yehuda, R., Golier, J. A., Halligan, S. L., Meaney, M., & Bierer, L. M. (2004).The ACTH response to dexamethasone in PTSD. American Journal of Psychiatry, 16, 1397-1403.

Yoshihama, M., Gillespie, B., Hammock, A. C., Belli, R. F., & Tolman, R. M. (2005). Does the Life History Calendar method facilitate the recall of intimate partner violence? Comparison of two methods of data collection. Social Work Research, 29, 151-163.

Zugazaga, C. (2004). Stressful life event experiences of homeless adults: A comparison of single men, single women, and women with children. Journal of Community Psychology, 32, 643-654.

Julie K. Williams, PhD, ACSW, LMSW, is codirector, Humility of Mary Housing, Inc. & Humility of Mary Shelter, Inc., 1228 East 12th Street, Davenport, IA 52803, and adjunct assistant professor, University of Iowa, Iowa City, IA 52241; e-mail: julie-k-williams@uiowa.edu. James A. Hall, PhD, LISW, is dean and professor, School of Social Work, University of Alabama, Tuscaloosa. This study was funded by the Social Sciences Funding Program through the University of Iowa. Address correspondence to Julie K. Williams, 1228 East 12th Street, Davenport, IA 52803.
Table 1: Multiple Regression Analysis of Level of Current
Traumatic Stress among Mothers Experiencing Homelessness

Variable                                    b         SE        t

Age                                       -0.06      0.40     -0.16
Months since current traumatic event      -0.06      0.30     -1.94
Number of past traumatic events            2.66      0.76      3.50
Average score of past traumatic events     2.43      1.49      1.62
Number of distressing events              -0.70      0.88     -0.80
Intercept                                 23.83     13.71      1.74

                                                       Confidence
                                                        Interval

Variable                                    p       Lower     Upper

Age                                       0.88      -0.87      0.74
Months since current traumatic event      0.06      -0.12      0.002
Number of past traumatic events           0.001      1.14      4.18
Average score of past traumatic events    0.11      -0.56      5.40
Number of distressing events              0.42      -2.45      1.04
Intercept                                 0.09      -3.52     51.18

Adjusted [R.sup.2] = .38, F(5, 69) = 10.18, p < .001
Gale Copyright: Copyright 2009 Gale, Cengage Learning. All rights reserved.